Daily Papers

Multimodal learning has revolutionized general domain tasks, yet its application in scientific discovery is hindered by the profound semantic gap between complex scientific imagery and sparse textual descriptions. We present S1-MMAlign, a large-scale, multi-disciplinary multimodal dataset comprising over 15.5 million high-quality image-text pairs derived from 2.5 million open-access scientific papers. Spanning disciplines from physics and biology to engineering, the dataset captures diverse visual modalities including experimental setups, heatmaps, and microscopic imagery. To address the pervasive issue of weak alignment in raw scientific captions, we introduce an AI-ready semantic enhancement pipeline that utilizes the Qwen-VL multimodal large model series to recaption images by synthesizing context from paper abstracts and citation contexts. Technical validation demonstrates that this enhancement significantly improves data quality: SciBERT-based pseudo-perplexity metrics show reduced semantic ambiguity, while CLIP scores indicate an 18.21% improvement in image-text alignment. S1-MMAlign provides a foundational resource for advancing scientific reasoning and cross-modal understanding in the era of AI for Science. The dataset is publicly available at https://huggingface.co/datasets/ScienceOne-AI/S1-MMAlign.

We present a multimodal dataset of 1020 hours of simultaneously recorded scalp electroencephalography (EEG), facial electromyography (EMG), and speech audio from three healthy native Japanese speakers during open-vocabulary overt speech. Recordings were acquired with three EEG systems-an ultra-high-density system (g.Pangolin) and two cap-type systems (g.SCARABEO and eegosports), spanning 62-128 channels-across many sessions over several months. Each session provides time-synchronized EEG, facial EMG, and audio, together with speech-event annotations and transcriptions. Although collected with speech decoding as a primary motivation, the dataset also supports work on multimodal signal processing, artifact modeling, longitudinal and cross-device adaptation, and EEG representation learning. Technical validation included power spectral density and event-related potential analyses across participants, devices, and tasks, which showed the expected 1/f spectral profile, task-related alpha-band attenuation, and time-locked evoked responses. The dataset is released in Brain Imaging Data Structure (BIDS) format via OpenNeuro under a CC0 waiver to support both speech-related and broader EEG research.

PlagiaCorpus is a bilingual Arabic-English dataset for benchmarking plagiarism and text-reuse detection, including reuse produced by large language models (LLMs) and reuse across languages. It contains 198,643 text records: a source corpus of 34,885 passages drawn from English and Arabic Wikipedia and two openly licensed book collections, and 163,758 queries (147,381 for training and 16,377 for testing). The queries cover five reuse classes of increasing difficulty, namely verbatim copying, character-level obfuscation, same-language paraphrasing by two different LLMs, and English--Arabic translation with structural paraphrasing, together with a Safe class of non-reused passages. Every positive query is linked to its source passage, and the train/test split is grouped by source passage so that no passage contributes queries to both splits. The release was curated through language identification, removal of LLM generation artefacts, and exact and cross-lingual de-duplication. Technical validation reports per-class length statistics, automatic checks of the train/test split, and inter-annotator agreement on a manually rated sample. The dataset is distributed in JSON and Parquet formats under a CC BY-SA 4.0 license.

Lower-Limb Fractures (LLF) are a major health concern for older adults, often leading to reduced mobility and prolonged recovery, potentially impairing daily activities and independence. During recovery, older adults frequently face social isolation and functional decline, complicating rehabilitation and adversely affecting physical and mental health. Multi-modal sensor platforms that continuously collect data and analyze it using machine-learning algorithms can remotely monitor this population and infer health outcomes. They can also alert clinicians to individuals at risk of isolation and decline. This paper presents a new publicly available multi-modal sensor dataset, MAISON-LLF, collected from older adults recovering from LLF in community settings. The dataset includes data from smartphone and smartwatch sensors, motion detectors, sleep-tracking mattresses, and clinical questionnaires on isolation and decline. The dataset was collected from ten older adults living alone at home for eight weeks each, totaling 560 days of 24-hour sensor data. For technical validation, supervised machine-learning and deep-learning models were developed using the sensor and clinical questionnaire data, providing a foundational comparison for the research community.

Speech-based analysis offers a scalable and non-invasive approach for detecting cognitive decline, yet progress has been constrained by the limited availability of clinically validated datasets collected under realistic conditions. We introduce PROCESS-2, a large-scale speech dataset designed to support research on automatic assessment of cognitive impairment from spontaneous and task-oriented speech. The dataset comprises recordings from 200 healthy controls, 150 mild cognitive impairment, and 50 dementia diagnoses collected using the CognoMemory digital assessment platform. Each participant completed a single assessment session, including picture description and verbal fluency tasks, accompanied by manually verified transcripts and participant-level metadata. PROCESS-2 contains approximately 21 hours of speech audio with predefined train/test partitions. Comprehensive technical validation evaluated demographic balance, clinical consistency, recording stability, embedding-space structure, and reproducible baseline modelling performance, demonstrating clinically meaningful group separation and stable performance across modelling approaches while preserving real-world conversational variability. PROCESS-2 is released under controlled access via Hugging Face to enable responsible reuse while protecting participant privacy, providing a reproducible benchmark resource for speech-based cognitive assessment research.

High-quality machine learning (ML)-ready datasets play a foundational role in developing new artificial intelligence (AI) models or fine-tuning existing models for scientific applications such as weather and climate analysis. Unfortunately, despite the growing development of new deep learning models for weather and climate, there is a scarcity of curated, pre-processed machine learning (ML)-ready datasets. Curating such high-quality datasets for developing new models is challenging particularly because the modality of the input data varies significantly for different downstream tasks addressing different atmospheric scales (spatial and temporal). Here we introduce WxC-Bench (Weather and Climate Bench), a multi-modal dataset designed to support the development of generalizable AI models for downstream use-cases in weather and climate research. WxC-Bench is designed as a dataset of datasets for developing ML-models for a complex weather and climate system, addressing selected downstream tasks as machine learning phenomenon. WxC-Bench encompasses several atmospheric processes from meso-beta (20 - 200 km) scale to synoptic scales (2500 km), such as aviation turbulence, hurricane intensity and track monitoring, weather analog search, gravity wave parameterization, and natural language report generation. We provide a comprehensive description of the dataset and also present a technical validation for baseline analysis. The dataset and code to prepare the ML-ready data have been made publicly available on Hugging Face -- https://huggingface.co/datasets/nasa-impact/WxC-Bench

Polyps in the colon are widely known cancer precursors identified by colonoscopy. Whilst most polyps are benign, the polyp's number, size and surface structure are linked to the risk of colon cancer. Several methods have been developed to automate polyp detection and segmentation. However, the main issue is that they are not tested rigorously on a large multicentre purpose-built dataset, one reason being the lack of a comprehensive public dataset. As a result, the developed methods may not generalise to different population datasets. To this extent, we have curated a dataset from six unique centres incorporating more than 300 patients. The dataset includes both single frame and sequence data with 3762 annotated polyp labels with precise delineation of polyp boundaries verified by six senior gastroenterologists. To our knowledge, this is the most comprehensive detection and pixel-level segmentation dataset (referred to as PolypGen) curated by a team of computational scientists and expert gastroenterologists. The paper provides insight into data construction and annotation strategies, quality assurance, and technical validation. Our dataset can be downloaded from https://doi.org/10.7303/syn26376615.

Real conversations between fraudsters and their targets are among the most informative artifacts for studying telephone scams, yet also the scarcest: passive honeypots overwhelmingly capture automated messages and hang-ups, large-scale studies characterize call metadata rather than dialogue, and manual scam-baiting does not scale. We present a dataset of real scam-call conversations collected by an active voice-agent honeypot. Dedicated numbers are seeded into the lead-generation channels fraud operations harvest; inbound callers are answered by a low-latency conversational agent that adopts a plausible target persona and sustains the interaction while every call is recorded, transcribed, and automatically labeled. Over an initial 53-day window we captured 10,015 inbound scam and spam calls (6,601 with two or more turns): roughly 895 hours of audio and 328,869 transcribed turns from 5,665 distinct originating numbers. Under a holistic classifier the substantive calls are predominantly predatory-but-legal lead generation ("spam", about three in five), while about one in seven is an outright "scam" (949 in this snapshot). Each call carries a turn-level transcript, three-channel audio, per-turn latency telemetry, and layers of automatic labels, including a holistic scam/spam/legitimate judgment corroborated by independent human review (75% agreement on the binary decision). We describe the collection system, the record structure, and technical validation of the corpus's realism and label quality, including that the agent is recognized as non-human in only about 5% of engaged calls. We also benchmark established scam-detection methods, where detectors trained on published synthetic dialogue collapse in precision on real traffic.

Short-video platforms show an increasing impact on people's daily lives nowadays, with billions of active users spending plenty of time each day. The interactions between users and online platforms give rise to many scientific problems across computational social science and artificial intelligence. However, despite the rapid development of short-video platforms, currently there are serious shortcomings in existing relevant datasets on three aspects: inadequate user-video feedback, limited user attributes and lack of video content. To address these problems, we provide a large-scale dataset with rich user behavior, attributes and video content from a real mobile short-video platform. This dataset covers 10,000 voluntary users and 153,561 videos, and we conduct four-fold technical validations of the dataset. First, we verify the richness of the behavior and attribute data. Second, we confirm the representing ability of the content features. Third, we provide benchmarking results on recommendation algorithms with our dataset. Finally, we explore the filter bubble phenomenon on the platform using the dataset. We believe the dataset could support the broad research community, including but not limited to user modeling, social science, human behavior understanding, etc. The dataset and code is available at https://github.com/tsinghua-fib-lab/ShortVideo_dataset.

There is a growing need for diverse, high-quality stuttered speech data, particularly in the context of Indian languages. This paper introduces Project Boli, a multi-lingual stuttered speech dataset designed to advance scientific understanding and technology development for individuals who stutter, particularly in India. The dataset constitutes (a) anonymized metadata (gender, age, country, mother tongue) and responses to a questionnaire about how stuttering affects their daily lives, (b) captures both read speech (using the Rainbow Passage) and spontaneous speech (through image description tasks) for each participant and (c) includes detailed annotations of five stutter types: blocks, prolongations, interjections, sound repetitions and word repetitions. We present a comprehensive analysis of the dataset, including the data collection procedure, experience summarization of people who stutter, severity assessment of stuttering events and technical validation of the collected data. The dataset is released as an open access to further speech technology development.

Scaling transformer language models creates an inherent tension between expressivity and memory efficiency. While unique weights across layers preserve functional specialization---from input-grounding to abstract refinement---they incur a substantial memory footprint. Conversely, standard depth-sharing enforces uniform transformations that collapse representational diversity and degrade modeling quality. We introduce Gated Recurrent Transformer, a recurrent depth transformer where fixed-depth prelude and coda blocks bracket a single shared core iterated R times. Inspired by gated recurrent neural networks, we employ a lightweight projection and an elementwise update gate---conditioned on the hidden state, the fixed prelude output, and noise resampled at every step---to modulate the recurrent update. This allows the model to specialize the input to the same few layers across recurrences, rather than requiring many unique layers to achieve functional diversity. Under an isoFLOPS constraint, a 3-layer Gated Recurrent Transformer matches the accuracy of a 12-layer GPT-2 Small baseline with similar training and inference FLOPs, and leads MoR and heavy-tail depth sampling in all nine scale-by-budget cells; at medium and large scale it approaches dense quality at the standard token budget and overtakes it at medium scale once that budget is doubled. Under an isoPARAMS constraint, deeper recurrence achieves a 2.76 validation loss versus 2.84 for a non-recurrent counterpart at matched parameter and data budget. Our results demonstrate that adaptive depth reuse is a principled strategy for trading parameters for quality: at large scale, 63% fewer parameters and 59% less peak decoding memory for a 10% increase in compiled generation latency.

The ongoing evolution of language models has led to the development of large-scale architectures that demonstrate exceptional performance across a wide range of tasks. However, these models come with significant computational and energy demands, as well as potential privacy implications. In this context, Small Reasoning Language Models (SRLMs) with approximately 0.5 billion parameters present a compelling alternative due to their remarkable computational efficiency and cost effectiveness, particularly in resource-constrained environments. Despite these advantages, the limited capacity of 0.5 billion parameter models poses challenges in handling complex tasks such as mathematical reasoning and code generation. This research investigates various training strategies, including supervised fine-tuning (SFT), knowledge distillation (KD), and reinforcement learning (RL), as well as their hybrid implementations, to enhance the performance of 0.5B SRLMs. We analyze effective methodologies to bridge the performance gap between SRLMS and larger models and present insights into optimal training pipelines tailored for these smaller architectures. Through extensive experimental validation and analysis, our work aims to provide actionable recommendations for maximizing the reasoning capabilities of 0.5B models.

Large Language Model (LLM) agents have developed rapidly in recent years to solve complex real-world problems using external tools. However, the scarcity of high-quality trajectories still hinders the development of stronger LLM agents. Most existing works on multi-turn dialogue synthesis validate correctness only at the trajectory level, which may overlook turn-level errors that can propagate during training and degrade model performance. To address these limitations, we introduce ToolMind, a large-scale, high-quality tool-agentic dataset with 160k synthetic data instances generated using over 20k tools and 200k augmented open-source data instances. Our data synthesis pipeline first constructs a function graph based on parameter correlations and then uses a multi-agent framework to simulate realistic user-assistant-tool interactions. Beyond trajectory-level validation, we employ fine-grained turn-level filtering to remove erroneous or suboptimal steps, ensuring that only high-quality reasoning traces are retained. This approach mitigates error amplification during training while preserving self-corrective reasoning signals essential for robust tool-use learning. Models fine-tuned on ToolMind show significant improvements over baselines on several benchmarks.

Site-specific fertilizer recommendation systems adapt nutrient advice to location, soil properties, crop type, and production targets, but scientific reuse is constrained when recommendation functions remain accessible mainly through interactive interfaces, outputs are not versioned, and trained approximations cannot be independently loaded or benchmarked. This technical report presents the Turba fertilizer machine learning stack, a three-layer open-source implementation for reproducible site-specific fertilizer recommendation in Morocco. turba-client provides programmatic access to publicly accessible site profiles, crop-specific target-yield spaces, and N, P_2O_5, and K_2O recommendation workflows; turba-data distributes analysis-ready snapshots; and turba-models packages crop-specific machine learning surrogates of recommendation outputs. The architecture links upstream retrieval, versioned analytical snapshots, reproducible cross-model benchmarking, and loadable offline surrogates while preserving the distinction between recommendation-system outputs, observed agricultural data, and model-generated predictions. The first dataset was constructed from 44,096 unique ESA WorldCereal locations. Scenario expansion across supported cereal workflows generated 132,017 crop-location recommendation requests under a medium target-yield setting. The resulting 22-variable dataset spans 10 regions, 66 provinces, and 1,149 communes. Nine regression families were evaluated under a fixed deterministic 80/20 protocol, and the current release packages five best-performing crop-specific models. The machine learning task is recommendation-function emulation rather than prediction of observed crop response. The stack provides a reproducible basis for spatial and temporal validation, uncertainty estimation, field-trial comparison, and future integration with additional data.

Identifying subtle technical errors within complex scientific and technical documents, especially those requiring multimodal interpretation (e.g., formulas in images), presents a significant hurdle for Large Language Models (LLMs) whose inherent error-correction tendencies can mask inaccuracies. This exploratory proof-of-concept (PoC) study investigates structured LLM context conditioning, informed by Persistent Workflow Prompting (PWP) principles, as a methodological strategy to modulate this LLM behavior at inference time. The approach is designed to enhance the reliability of readily available, general-purpose LLMs (specifically Gemini 2.5 Pro and ChatGPT Plus o3) for precise validation tasks, crucially relying only on their standard chat interfaces without API access or model modifications. To explore this methodology, we focused on validating chemical formulas within a single, complex test paper with known textual and image-based errors. Several prompting strategies were evaluated: while basic prompts proved unreliable, an approach adapting PWP structures to rigorously condition the LLM's analytical mindset appeared to improve textual error identification with both models. Notably, this method also guided Gemini 2.5 Pro to repeatedly identify a subtle image-based formula error previously overlooked during manual review, a task where ChatGPT Plus o3 failed in our tests. These preliminary findings highlight specific LLM operational modes that impede detail-oriented validation and suggest that PWP-informed context conditioning offers a promising and highly accessible technique for developing more robust LLM-driven analytical workflows, particularly for tasks requiring meticulous error detection in scientific and technical documents. Extensive validation beyond this limited PoC is necessary to ascertain broader applicability.

We present a modular semantic account of Bayesian inference algorithms for probabilistic programming languages, as used in data science and machine learning. Sophisticated inference algorithms are often explained in terms of composition of smaller parts. However, neither their theoretical justification nor their implementation reflects this modularity. We show how to conceptualise and analyse such inference algorithms as manipulating intermediate representations of probabilistic programs using higher-order functions and inductive types, and their denotational semantics. Semantic accounts of continuous distributions use measurable spaces. However, our use of higher-order functions presents a substantial technical difficulty: it is impossible to define a measurable space structure over the collection of measurable functions between arbitrary measurable spaces that is compatible with standard operations on those functions, such as function application. We overcome this difficulty using quasi-Borel spaces, a recently proposed mathematical structure that supports both function spaces and continuous distributions. We define a class of semantic structures for representing probabilistic programs, and semantic validity criteria for transformations of these representations in terms of distribution preservation. We develop a collection of building blocks for composing representations. We use these building blocks to validate common inference algorithms such as Sequential Monte Carlo and Markov Chain Monte Carlo. To emphasize the connection between the semantic manipulation and its traditional measure theoretic origins, we use Kock's synthetic measure theory. We demonstrate its usefulness by proving a quasi-Borel counterpart to the Metropolis-Hastings-Green theorem.

We introduce SAGE Celer 2.6, the latest in our line of general-purpose Celer models from SAGEA. Celer 2.6 is available in 5B, 10B, and 27B parameter sizes and benefits from extensive architectural modifications and further pre-training on an undisclosed model. Using our Inverse Reasoning (IR) pipeline, SAGEA natively trains Celer 2.6 to validate its own logic paths, minimizing cascading error and hallucination in complex reasoning tasks. Celer 2.6 also boasts natively integrated multimodal functionality with an end-to-end vision encoder to avoid common pitfalls in adapter-based approaches. Celer 2.6 provides highly competitive results on mathematics, coding, and general intelligence benchmarks (ACUMEN), along with low latency. Most importantly, Celer 2.6 is specifically optimized for South Asian language support, with a custom tokenizer for the Devanagari script and strong performance in both Nepali and Hindi without sacrificing English reasoning ability.

Large language models (LLMs) have revolutionized artificial intelligence by enabling complex reasoning capabilities. While recent advancements in reinforcement learning (RL) have primarily focused on domain-specific reasoning tasks (e.g., mathematics or code generation), real-world reasoning scenarios often require models to handle diverse and complex environments that narrow-domain benchmarks cannot fully capture. To address this gap, we present InternBootcamp, an open-source framework comprising 1000+ domain-diverse task environments specifically designed for LLM reasoning research. Our codebase offers two key functionalities: (1) automated generation of unlimited training/testing cases with configurable difficulty levels, and (2) integrated verification modules for objective response evaluation. These features make InternBootcamp fundamental infrastructure for RL-based model optimization, synthetic data generation, and model evaluation. Although manually developing such a framework with enormous task coverage is extremely cumbersome, we accelerate the development procedure through an automated agent workflow supplemented by manual validation protocols, which enables the task scope to expand rapidly. % With these bootcamps, we further establish Bootcamp-EVAL, an automatically generated benchmark for comprehensive performance assessment. Evaluation reveals that frontier models still underperform in many reasoning tasks, while training with InternBootcamp provides an effective way to significantly improve performance, leading to our 32B model that achieves state-of-the-art results on Bootcamp-EVAL and excels on other established benchmarks. In particular, we validate that consistent performance gains come from including more training tasks, namely task scaling, over two orders of magnitude, offering a promising route towards capable reasoning generalist.

We present the GEST-Engine, a complete system that goes from natural-language text to fully-annotated multi-actor video. At its core is an explicit world model: rather than encoding state as a learned latent, the engine maintains a complete, inspectable representation of the world (which actors exist, where they are, what they are doing, which objects they hold, and how events relate in time and space), expressed as a formal Graph of Events in Space and Time (GEST) and realized deterministically inside the open world of a commercial game engine driven through an open-source multiplayer scripting framework. GESTs are produced either procedurally or by an agentic text-to-GEST system in which an LLM Director plans a story through tool calls validated by a programmatic state backend, so every generated specification is executable by construction. A GEST then enters a four-stage execution pipeline: graph parsing and validation, entity and action grounding, temporal orchestration (Allen-style constraints resolved by Floyd-Warshall transitive closure), and execution and capture. In a single simulation pass the engine emits frame-aligned RGB video, dense per-pixel depth, instance segmentation, per-actor skeletal pose, per-frame pairwise spatial-relation graphs, 2D bounding boxes, event-to-frame temporal mappings, and natural-language descriptions, all at zero marginal annotation cost. We further describe an in-game world editor, runtime capability extraction, a text-generation pipeline, and a production system that renders corpora at scale across parallel virtual machines. Because every frame traces back to a semantic specification, the engine guarantees object permanence, multi-actor coordination, and temporal consistency by construction, making its output valuable as training data, evaluation benchmarks, and diagnostic tools for video understanding.

Despite progresses in data engineering, there are areas with limited consistencies across data validation and documentation procedures causing confusions and technical problems in research involving machine learning. There have been progresses by introducing frameworks like "Datasheets for Datasets", however there are areas for improvements to prepare datasets, ready for ML pipelines. Here, we extend the framework to "Datasheets for AI and medical datasets - DAIMS." Our publicly available solution, DAIMS, provides a checklist including data standardization requirements, a software tool to assist the process of the data preparation, an extended form for data documentation and pose research questions, a table as data dictionary, and a flowchart to suggest ML analyses to address the research questions. The checklist consists of 24 common data standardization requirements, where the tool checks and validate a subset of them. In addition, we provided a flowchart mapping research questions to suggested ML methods. DAIMS can serve as a reference for standardizing datasets and a roadmap for researchers aiming to apply effective ML techniques in their medical research endeavors. DAIMS is available on GitHub and as an online app to automate key aspects of dataset evaluation, facilitating efficient preparation of datasets for ML studies.

Neuro-endoscopy is a challenging minimally invasive neurosurgery that requires surgical skills to be acquired using training methods different from the existing apprenticeship model. There are various training systems developed for imparting fundamental technical skills in laparoscopy where as limited systems for neuro-endoscopy. Neuro-Endo-Trainer was a box-trainer developed for endo-nasal transsphenoidal surgical skills training with video based offline evaluation system. The objective of the current study was to develop a modified version (Neuro-Endo-Trainer-Online Assessment System (NET-OAS)) by providing a stand-alone system with online evaluation and real-time feedback. The validation study on a group of 15 novice participants shows the improvement in the technical skills for handling the neuro-endoscope and the tool while performing pick and place activity.

Large-scale clinical databases offer opportunities for medical research, but their complexity creates barriers to effective use. The Medical Information Mart for Intensive Care (MIMIC-IV), one of the world's largest open-source electronic health record databases, traditionally requires both SQL proficiency and clinical domain expertise. We introduce M3, a system that enables natural language querying of MIMIC-IV data through the Model Context Protocol. With a single command, M3 retrieves MIMIC-IV from PhysioNet, launches a local SQLite instance or connects to hosted BigQuery, and allows researchers to pose clinical questions in plain English. We evaluated M3 using one hundred questions from the EHRSQL 2024 benchmark with two language models: the proprietary Claude Sonnet 4 achieved 94% accuracy, while the open-source gpt-oss-20B (deployable locally on consumer hardware) achieved 93% accuracy. Both models translate natural language into SQL, execute queries against MIMIC-IV, and return structured results alongside the underlying query for verification. Error analysis revealed that most failures stemmed from complex temporal reasoning or ambiguous question phrasing rather than fundamental architectural limitations. The comparable performance of a smaller open-source model demonstrates that privacy-preserving local deployment is viable for sensitive clinical data analysis. M3 lowers technical barriers to critical care data analysis while maintaining security through OAuth2 authentication, query validation, and comprehensive audit logging.

The deliberate manipulation of public opinion, especially through altered images, which are frequently disseminated through online social networks, poses a significant danger to society. To fight this issue on a technical level we support the research community by releasing the Digital Forensics 2023 (DF2023) training and validation dataset, comprising one million images from four major forgery categories: splicing, copy-move, enhancement and removal. This dataset enables an objective comparison of network architectures and can significantly reduce the time and effort of researchers preparing datasets.

Environmental, Social, and Governance (ESG) standards have been increasingly adopted by organizations to demonstrate accountability towards ethical, social, and sustainability goals. However, generating ESG reports that align with these standards remains challenging due to unstructured data formats, inconsistent terminology, and complex requirements. Existing ESG lifecycles provide guidance for structuring ESG reports but lack the automation, adaptability, and continuous feedback mechanisms needed to address these challenges. To bridge this gap, we introduce an agentic ESG lifecycle framework that systematically integrates the ESG stages of identification, measurement, reporting, engagement, and improvement. In this framework, multiple AI agents extract ESG information, verify ESG performance, and update ESG reports based on organisational outcomes. By embedding agentic components within the ESG lifecycle, the proposed framework transforms ESG from a static reporting process into a dynamic, accountable, and adaptive system for sustainability governance. We further define the technical requirements and quality attributes needed to support four main ESG tasks, such as report validation, multi-report comparison, report generation, and knowledge-base maintenance, and propose three architectural approaches, namely single-model, single-agent, and multi-agent, for addressing these tasks. The source code and data for the prototype of these approaches are available at https://gitlab.com/for_peer_review-group/esg_assistant.

Large language models and multilingual machine translation (MT) systems increasingly drive access to information, yet many languages of the tribal communities remain effectively invisible in these technologies. This invisibility exacerbates existing structural inequities in education, governance, and digital participation. We present AdiBhashaa, a community-driven initiative that constructs the first open parallel corpora and baseline MT systems for four major Indian tribal languages-Bhili, Mundari, Gondi, and Santali. This work combines participatory data creation with native speakers, human-in-the-loop validation, and systematic evaluation of both encoder-decoder MT models and large language models. In addition to reporting technical findings, we articulate how AdiBhashaa illustrates a possible model for more equitable AI research: it centers local expertise, builds capacity among early-career researchers from marginalized communities, and foregrounds human validation in the development of language technologies.

Generative artificial intelligence (AI) systems open new possibilities for experimentation in literary studies via controlled, grounded, large-scale, low-cost simulations of cultural production. Current systems have not yet been shown to produce high-quality, book-length narrative texts that reliably reflect arbitrarily specified cultural constraints or stylistic features. But there exists substantial relevant research on each of the components required for literary-historical simulation. These include the use and validation of AI systems as proxies for differentiable human populations; the narrative and stylistic properties of AI-generated texts; the stability and coherence of multiagent, multiturn AI simulations of human actors; and technical methods through which to alter in predictable ways the knowledge and behavior of generative systems. Together, these areas could provide a starting point for more ambitious AI-based modeling of cultural systems of literary production. We describe the possibilities and challenges of simulation-based experiments in literary studies, summarize the current state of the art in relevant fields, and explain key technical aspects of the work. To provide an example directly relevant to literary scholars, we present the results of experiments on literary text generation, including comparisons to high-status, human-authored novels. Our results include the first demonstration of (limited) in-distribution outputs by AI models in this domain. We conclude with a description of future work on full counterfactual literary-historical simulations using AI.

Current video generation models produce high-quality aesthetic videos but often struggle to learn representations of real-world physics dynamics, resulting in artifacts such as unnatural object collisions, inconsistent gravity, and temporal flickering. In this work, we propose PhysVideoGenerator, a proof-of-concept framework that explicitly embeds a learnable physics prior into the video generation process. We introduce a lightweight predictor network, PredictorP, which regresses high-level physical features extracted from a pre-trained Video Joint Embedding Predictive Architecture (V-JEPA 2) directly from noisy diffusion latents. These predicted physics tokens are injected into the temporal attention layers of a DiT-based generator (Latte) via a dedicated cross-attention mechanism. Our primary contribution is demonstrating the technical feasibility of this joint training paradigm: we show that diffusion latents contain sufficient information to recover V-JEPA 2 physical representations, and that multi-task optimization remains stable over training. This report documents the architectural design, technical challenges, and validation of training stability, establishing a foundation for future large-scale evaluation of physics-aware generative models.

Large language models (LLMs) are increasingly deployed across healthcare applications, including clinical documentation, diagnostic reasoning, medicine recommendation, and medical education. Their outputs are largely unstructured clinical text, which is difficult to reliably evaluate at scale. LLM-as-a-Judge, in which an LLM evaluates another system's output against task-specific criteria, offers a scalable alternative and is increasingly used in clinical evaluation, yet its validity in healthcare remains underexamined. Existing reviews focus on general-purpose LLM evaluation or on risk framework, rather than systematically characterizing how LLM-as-a-Judge is applied in healthcare and how well their judgments align with human experts. We therefore conduct a PRISMA-guided comprehensive review of LLM-as-a-Judge applications in healthcare, searching five databases for studies published between January 2023 and February 2026. After screening 541 records, 134 studies meet the eligibility and are coded by health scenario, judge configuration, technical approach, and validation design. LLM-as-a-Judge is concentrated in clinical decision support, clinical natural language processing (NLP), medical knowledge and question answering (QA), and medical communication. OpenAI models are the most frequently used judges, and prompt engineering appears in nearly all studies, with ensemble, multi-agent, and retrieval-augmented designs as common extensions. Among studies reporting human validation, LLM judges often show moderate to strong alignment with expert judgments, although reliability varies substantially across tasks. Overall, this review positions LLM-as-a-Judge as a promising framework for scalable healthcare AI evaluation, while emphasizing that its clinical value depends on model design and rigorous validation.

In industrial procurement, an LLM answer is useful only if it survives a standards check: recommended material must match operating condition, every parameter must respect a regulated threshold, and no procedure may contradict a safety clause. Partial correctness can mask safety-critical contradictions that aggregate LLM benchmarks rarely capture. We introduce IndustryBench, a 2,049-item benchmark for industrial procurement QA in Chinese, grounded in Chinese national standards (GB/T) and structured industrial product records, organized by seven capability dimensions, ten industry categories, and panel-derived difficulty tiers, with item-aligned English, Russian, and Vietnamese renderings. Our construction pipeline rejects 70.3% of LLM-generated candidates at a search-based external-verification stage, calibrating how unreliable industrial QA remains after LLM-only filtering.Our evaluation decouples raw correctness, scored by a Qwen3-Max judge validated at κ_w = 0.798 against a domain expert, from a separate safety-violation (SV) check against source texts. Across 17 models in Chinese and an 8-model intersection over four languages, we find: (i) the best system reaches only 2.083 on the 0--3 rubric, leaving substantial headroom; (ii) Standards & Terminology is the most persistent capability weakness and survives item-aligned translation; (iii) extended reasoning lowers safety-adjusted scores for 12 of 13 models, primarily by introducing unsupported safety-critical details into longer final answers; and (iv) safety-violation rates reshuffle the leaderboard -- GPT-5.4 climbs from rank 6 to rank 3 after SV adjustment, while Kimi-k2.5-1T-A32B drops seven positions.Industrial LLM evaluation therefore requires source-grounded, safety-aware diagnosis rather than aggregate accuracy. We release IndustryBench with all prompts, scoring scripts, and dataset documentation.

We introduce TechQA, a domain-adaptation question answering dataset for the technical support domain. The TechQA corpus highlights two real-world issues from the automated customer support domain. First, it contains actual questions posed by users on a technical forum, rather than questions generated specifically for a competition or a task. Second, it has a real-world size -- 600 training, 310 dev, and 490 evaluation question/answer pairs -- thus reflecting the cost of creating large labeled datasets with actual data. Consequently, TechQA is meant to stimulate research in domain adaptation rather than being a resource to build QA systems from scratch. The dataset was obtained by crawling the IBM Developer and IBM DeveloperWorks forums for questions with accepted answers that appear in a published IBM Technote---a technical document that addresses a specific technical issue. We also release a collection of the 801,998 publicly available Technotes as of April 4, 2019 as a companion resource that might be used for pretraining, to learn representations of the IT domain language.

Technical standards, or simply standards, are established documented guidelines and rules that facilitate the interoperability, quality, and accuracy of systems and processes. In recent years, we have witnessed an emerging paradigm shift where the adoption of generative AI (GenAI) models has increased tremendously, spreading implementation interests across standard-driven industries, including engineering, legal, healthcare, and education. In this paper, we assess the criticality levels of different standards across domains and sectors and complement them by grading the current compliance capabilities of state-of-the-art GenAI models. To support the discussion, we outline possible challenges and opportunities with integrating GenAI for standard compliance tasks while also providing actionable recommendations for entities involved with developing and using standards. Overall, we argue that aligning GenAI with standards through computational methods can help strengthen regulatory and operational compliance. We anticipate this area of research will play a central role in the management, oversight, and trustworthiness of larger, more powerful GenAI-based systems in the near future.

We consider the task of mapping pseudocode to long programs that are functionally correct. Given test cases as a mechanism to validate programs, we search over the space of possible translations of the pseudocode to find a program that passes the validation. However, without proper credit assignment to localize the sources of program failures, it is difficult to guide search toward more promising programs. We propose to perform credit assignment based on signals from compilation errors, which constitute 88.7% of program failures. Concretely, we treat the translation of each pseudocode line as a discrete portion of the program, and whenever a synthesized program fails to compile, an error localization method tries to identify the portion of the program responsible for the failure. We then focus search over alternative translations of the pseudocode for those portions. For evaluation, we collected the SPoC dataset (Search-based Pseudocode to Code) containing 18,356 programs with human-authored pseudocode and test cases. Under a budget of 100 program compilations, performing search improves the synthesis success rate over using the top-one translation of the pseudocode from 25.6% to 44.7%.

This position paper argues that the machine learning community should prioritize early-stage quality assurance in annotation pipelines over the prevailing practice of late-stage validation. Data quality bottlenecks increasingly limit foundation model improvement, yet quality assurance research focuses almost exclusively on validation methods rather than validation timing. When validation occurs, not merely what methods are employed, fundamentally determines both error rates and annotation costs. This temporal neglect is puzzling given the well-established "shift-left" principle from software engineering, where empirical studies demonstrate 4--100x cost multipliers for defects detected in later stages (Boehm, 1981; Shull et al., 2002). Annotation pipelines exhibit analogous dynamics: errors caught before annotation begins cost a fraction of those discovered after review cycles complete. We propose a taxonomy of three QA trigger points, namely pre-annotation (T0), post-annotation (T1), and post-review (T2), that decompose annotation workflows into discrete validation opportunities. A parametric error-propagation model formalizes when timing affects final error rates versus only economics, making timing a measurable design variable rather than a configuration afterthought. A survey of 47 recent papers reveals that only 4% report when validation occurs, a striking gap given timing's demonstrated impact in adjacent fields. Without explicit attention to QA timing, the community risks optimizing validation methods while ignoring the structural variable that may matter most. Acting on this position requires three steps: researchers should report QA timing configurations alongside validation methods; annotation platforms should expose timing as a first-class parameter; and the community should run controlled experiments that measure stage-specific detection rates directly.

Self-driving laboratories couple a learned design surrogate to an instrument, and on every new target domain it must answer four questions: when it can be trusted, how many target measurements it needs, which to make first, and when to stop adapting, switch models or escalate. We state a framework making each measurable (a competence ratio, a budget k^*(varepsilon), a time-to-target and a priced decision) and instantiate it without wet experiments on six public droplet-microfluidic and liposome tables as offline laboratories: 107 leave-one-group tiers whose rows are hidden labels. A sealed, pre-registered programme (two sealed negative validations and a pre-registered test on three more tables) and seven frozen closed-loop protocols, each run twice, yield 23 findings. Shift cost is stratified by shift type; no label-free statistic certifies individual candidates; random-split model-family rankings predicted neither the sealed nor the loop rankings. Warm-start residual boosting returns a surrogate to 1.25times its development reference within 20 target rows in 94% of chemistry tiers (27% under a strict noise-floor reading), 90% of laboratory and 40% of device tiers, and a budget planned from the shift type on other tables is not measurably worse on a new one. In the loop, one gain survives drift and shuffled-label controls on all tiers, carried by the sparsely sampled ones: proposing at once and warm-starting on every measurement reaches a 10% target in 2.80 measurements against 3.31 for the reference, and in 5.72 against 11.35 for random proposals on 301 hard units at a 3% tolerance. No label-free acquisition rule survives the shuffled-label control, no interval-aware proposal or other model family does better, and no threshold rule on the tested diagnostics beats always continuing once escalation is priced. No wet-lab validation is claimed.

We introduce PHYSICS, a comprehensive benchmark for university-level physics problem solving. It contains 1297 expert-annotated problems covering six core areas: classical mechanics, quantum mechanics, thermodynamics and statistical mechanics, electromagnetism, atomic physics, and optics. Each problem requires advanced physics knowledge and mathematical reasoning. We develop a robust automated evaluation system for precise and reliable validation. Our evaluation of leading foundation models reveals substantial limitations. Even the most advanced model, o3-mini, achieves only 59.9% accuracy, highlighting significant challenges in solving high-level scientific problems. Through comprehensive error analysis, exploration of diverse prompting strategies, and Retrieval-Augmented Generation (RAG)-based knowledge augmentation, we identify key areas for improvement, laying the foundation for future advancements.

For scientific progress, we need benchmarks that test the limits of state-of-the-art models, and evaluation methods that inform us about failure cases. As models get stronger, standard benchmarks for machine translation are approaching saturation. Further, automatic translation metrics are unreliable, vulnerable to reward-hacking, and provide unactionable assessments. Even gold human evaluation is not problem-free, because it often lacks reproducibility, objectivity, and scalability. Overall, this prevents us from tracking objective progress in the field and identifying pathways for improvement. We introduce the Last Translation Benchmark, a collection of human-authored and peer-reviewed examples (texts, images, audio, videos) that break leading machine translation models. We also present a new evaluation approach: each example comes with handcrafted verification rules describing concrete failure cases on that example, therefore allowing reliable and actionable future evaluation. The Last Translation Benchmark is a live dataset that accepts ongoing contributions. The latest version is LTBv1, containing accepted contributions prior to September 1st 2026, with future releases planned as new data is continuously collected.

Codes of professional ethics guide and restrict specialists, who increasingly work with assistants built on large language models. Such an assistant can breach the user's code, or help the user breach it, without the user noticing. Evaluations reporting broad categories of harm cannot show how often this happens or which provision is breached. TenetBench pairs 3,433 requests conflicting with clauses of fifteen Russian-language professional codes with benign twins that differ only in the legitimacy of what is asked. An automated pipeline generates the tasks from the text of a code, so the corpus grows with every code added and a provision yields new tasks on every run. It also produced tasks and human-validated verdicts from two English codes of other jurisdictions. In human validation of the frozen pipeline, specialists rated 0.995 of the conflicting tasks plausible, and the model judge agreed with the adjudicated label at Gwet's AC1 of at least 0.894 on 1,860 elements. We evaluated fourteen model and reasoning-mode combinations on whether they produce the conflicting text, decline or warn. Eleven of the fourteen produce conflicting content on most such requests, mostly without a warning, and nearly three warnings in four accompany conflicting content. The models that conform more do so by declining while still serving the legitimate twin, and enabling reasoning halved the refusals of two models. We hope the dataset and pipeline will help evaluate how language models observe professional norms and let diagnostic datasets be built for particular professions and organisations.

Large language models (LLMs) can be leveraged to help with writing formulas in spreadsheets, but resources on these formulas are scarce, impacting both the base performance of pre-trained models and limiting the ability to fine-tune them. Given a corpus of formulas, we can use a(nother) model to generate synthetic natural language utterances for fine-tuning. However, it is important to validate whether the NL generated by the LLM is indeed accurate to be beneficial for fine-tuning. In this paper, we provide empirical results on the impact of validating these synthetic training examples with surrogate objectives that evaluate the accuracy of the synthetic annotations. We demonstrate that validation improves performance over raw data across four models (2 open and 2 closed weight). Interestingly, we show that although validation tends to prune more challenging examples, it increases the complexity of problems that models can solve after being fine-tuned on validated data.

Agents need reliable ways to validate their outputs, both when completing tasks and when filtering synthetic training data. We study verifier generation: from the instruction and environment available for solving a task, an agent writes an executable verifier that must accept every valid solution and reject invalid ones. We introduce VerifierBench, a benchmark of 222 tasks from six agent benchmarks spanning tool-call, state-based, and code verification, with 4,993 hidden test cases that include alternative valid solutions and attacks; Verifier4Verifier (V4V), a six-check audit, screens its source verifiers and test suites. Across 19 models in a common agent harness, normalized balanced accuracy ranges from 10.6 to 62.1, with generated verifiers consistently rejecting invalid solutions more reliably than they accept valid ones. This apparent strength can mask overly restrictive verification: most false rejections in the analyzed records come from verifiers that accept no tested solution to a task. Case studies reveal unsupported assumptions about solution formats, tool use, and code changes that exclude valid solutions. These errors survive self-testing because agents test solutions that share the verifier's mistaken assumptions, or submit verifiers without confirming that any correct solution passes. These findings suggest testing both whether a verifier accepts a correct solution and whether it accepts other solutions permitted by the task. The project page and dataset are publicly available.

Coding agents are increasingly asked to build software rather than patch it, yet benchmarks for from-scratch repository construction are mostly limited to a single language and depend on manually curated tasks. We introduce Zero2Repo, a benchmark in which an agent receives a product requirements document, an interface contract, and an empty workspace, and must deliver a complete repository in the project's native ecosystem. Tasks are produced by a language-agnostic authoring pipeline that converts real, version-pinned open-source projects into behavioral specifications, reproducible environments, and hidden acceptance tests. Each task is validated by execution: a reference implementation derived from the upstream project must pass, and adversarial validation must show that the tests reject incorrect implementations. Evaluation runs production coding agents in isolated containers, withholds the acceptance tests until an explicit submission, and assigns a binary reward only when every test passes, with no LLM judge. The pipeline and harness make no language-specific assumptions and apply to mainstream programming ecosystems; the current release contains Python, TypeScript, Go, and C++ tasks. Even on 11 tasks drawn from repositories that frontier models have very likely seen during training, the strongest agent solves only 10, and every failing submission passes 90-99% of the hidden tests; for the two strongest agents, 67-100% of failed tests trace to a single omission or a low-frequency rule stated in the specification rather than to a missing subsystem, so each failure is a concrete target for improvement.

Automated code generation is becoming an important tool in quantitative finance, where large language models can generate option pricing implementations directly from mathematical model specifications. Validating such implementations, however, requires considerably more than conventional software testing: numerical pricing methods must remain mathematically consistent, numerically stable, and reliable across a wide range of model parameters. We introduce RIDGE, an autonomous validation framework in which generated pricing implementations are subjected to structured no-arbitrage tests, stress tests, benchmark comparisons, and consistency checks. Validation evidence is interpreted diagnostically, while the resulting knowledge is accumulated in a repository and reused across models and successive validation iterations. This enables systematic refinement of both the pricing implementation and the validation methodology. The framework is applied to five stochastic volatility models. Across these studies, all detected implementation defects are removed and, in two cases, the validation process itself leads to new semi-analytic pricing methodologies. The supplementary material is available in the GitHub repository: https://github.com/ShQiangLiu/ridge.

Large language models are increasingly used to generate code from natural language, but ensuring correctness remains challenging. Formal verification offers a principled way to obtain such guarantees by proving that a program satisfies a formal specification. However, specifications are frequently missing in real-world codebases, and writing high-quality specifications remains expensive and expertise-intensive. We present VeriSpecGen, a traceable refinement framework that synthesizes intent-aligned specifications in Lean through requirement-level attribution and localized repair. VeriSpecGen decomposes natural language into atomic requirements and generates requirement-targeted tests with explicit traceability maps to validate generated specifications. When validation fails, traceability maps attribute failures to specific requirements, enabling targeted clause-level repairs. VeriSpecGen achieve 86.6% on VERINA SpecGen task using Claude Opus 4.5, improving over baselines by up to 31.8 points across different model families and scales. Beyond inference-time gains, we generate 343K training examples from VeriSpecGen refinement trajectories and demonstrate that training on these trajectories substantially improves specification synthesis by 62-106% relative and transfers gains to general reasoning abilities.

Recent advances in Large Language Models (LLMs) have sparked growing interest in applying them to Electronic Design Automation (EDA) tasks, particularly Register Transfer Level (RTL) code generation. While several RTL datasets have been introduced, most focus on syntactic validity rather than functional validation with tests, leading to training examples that compile but may not implement the intended behavior. We present VERICODER, a model for RTL code generation fine-tuned on a dataset validated for functional correctness. This fine-tuning dataset is constructed using a novel methodology that combines unit test generation with feedback-directed refinement. Given a natural language specification and an initial RTL design, we prompt a teacher model (GPT-4o-mini) to generate unit tests and iteratively revise the RTL design based on its simulation results using the generated tests. If necessary, the teacher model also updates the tests to ensure they comply with the natural language specification. As a result of this process, every example in our dataset is functionally validated, consisting of a natural language description, an RTL implementation, and passing tests. Fine-tuned on this dataset of over 125,000 examples, VERICODER achieves state-of-the-art metrics in functional correctness on VerilogEval and RTLLM, with relative gains of up to 71.7% and 27.4% respectively. An ablation study further shows that models trained on our functionally validated dataset outperform those trained on functionally non-validated datasets, underscoring the importance of high-quality datasets in RTL code generation.

Large Language Models (LLMs) have exploded a new heatwave of AI, for their ability to engage end-users in human-level conversations with detailed and articulate answers across many knowledge domains. In response to their fast adoption in many industrial applications, this survey concerns their safety and trustworthiness. First, we review known vulnerabilities of the LLMs, categorising them into inherent issues, intended attacks, and unintended bugs. Then, we consider if and how the Verification and Validation (V&V) techniques, which have been widely developed for traditional software and deep learning models such as convolutional neural networks, can be integrated and further extended throughout the lifecycle of the LLMs to provide rigorous analysis to the safety and trustworthiness of LLMs and their applications. Specifically, we consider four complementary techniques: falsification and evaluation, verification, runtime monitoring, and ethical use. Considering the fast development of LLMs, this survey does not intend to be complete (although it includes 300 references), especially when it comes to the applications of LLMs in various domains, but rather a collection of organised literature reviews and discussions to support the quick understanding of the safety and trustworthiness issues from the perspective of V&V.

Large language models (LLMs) and vision-language models (VLMs) are usually frozen after deployment, so they do not learn from the cases they solve. This is especially concerning in medicine, where new clinical evidence, updated guidelines, and new therapies can change established practice. Fine-tuning can update the model, but it requires access to model weights and additional training. Parameter-free methods avoid training, but they may overfit a fixed validation set, lack reliable domain knowledge, or lose visual details by saving experience only as text. To address these limitations, we present a model-agnostic framework that allows frozen LLMs and VLMs to learn from deployment experience through three forms of external expertise: a Skill that guides reasoning and tool use, a Knowledge Memory that stores reliable facts supported by earlier cases or trusted external evidence, and a Multimodal Knowledge Base that keeps visual examples and guides the model to relate each retrieved case to the current image. Instead of relying on a fixed validation set, a validation strategy keeps an update only if it helps on new cases without degrading performance on earlier ones. Across six benchmarks covering clinical diagnosis, clinical workflows, medical reasoning, and medical and non-medical visual reasoning, and with four open-weight and closed-source base models, our framework improves performance during online deployment by up to 34.2% over the base model on medical tasks, generalizes to unseen cases, transfers to other models without further optimization, and works in non-medical domains.

SciCode is the standard measure of the scientific-coding ability of language models: research-level problems that demand both frontier scientific theory and its implementation as working numerical code. It is a component of the Artificial Analysis Intelligence Index and a standing evaluation in government and national-laboratory suites. Yet its scores have recently plateaued: the strongest 2026 models cluster tightly around 60\% subproblem accuracy, and a successor model ties its predecessor. We trace this stagnation to defects in the benchmark itself. A per-problem, domain-expert audit of all 65 test problems uncovers 263 defects; 192 of them, spread across 91\% of the main problems, cause correct, instruction-following solutions to be wrongly rejected---through non-reproducible gold answers, over-tight tolerances, or self-contradictory specifications. Critically, 78\% of these score-suppressing defects require specialized physics or mathematics knowledge to detect, not mere clerical proofreading. We corrected every confirmable defect to produce SciCode-Verified. The corrections add only the specifications a well-posed problem requires, repair grading, and tighten the tests that were too lenient; every change is recorded with its justification and independently re-checked by a second domain expert. We re-evaluate twelve frontier model snapshots on the corrected benchmark and find a substantial recovery: subproblem accuracy rises from 45--60\% to 84--98\%, and main-problem accuracy from 9--27\% to 69--92\%. State-of-the-art models are far more proficient in scientific coding than SciCode has suggested---the bottleneck was not model capability, but the quality of the evaluation instrument. We release SciCode-Verified with its complete audit trail as the corrected public standard.

Humanity's Last Exam (HLE) has become a widely used benchmark for evaluating frontier large language models on challenging, multi-domain questions. However, community-led analyses have raised concerns that HLE contains a non-trivial number of noisy items, which can bias evaluation results and distort cross-model comparisons. To address this challenge, we introduce HLE-Verified, a verified and revised version of HLE with a transparent verification protocol and fine-grained error taxonomy. Our construction follows a two-stage validation-and-repair workflow resulting in a certified benchmark. In Stage I, each item undergoes binary validation of the problem and final answer through domain-expert review and model-based cross-checks, yielding 641 verified items. In Stage II, flawed but fixable items are revised under strict constraints preserving the original evaluation intent, through dual independent expert repairs, model-assisted auditing, and final adjudication, resulting in 1,170 revised-and-certified items. The remaining 689 items are released as a documented uncertain set with explicit uncertainty sources and expertise tags for future refinement. We evaluate seven state-of-the-art language models on HLE and HLE-Verified, observing an average absolute accuracy gain of 7--10 percentage points on HLE-Verified. The improvement is particularly pronounced on items where the original problem statement and/or reference answer is erroneous, with gains of 30--40 percentage points. Our analyses further reveal a strong association between model confidence and the presence of errors in the problem statement or reference answer, supporting the effectiveness of our revisions. Overall, HLE-Verified improves HLE-style evaluations by reducing annotation noise and enabling more faithful measurement of model capabilities. Data is available at: https://github.com/SKYLENAGE-AI/HLE-Verified

Given a black-box AI system and a task, at what confidence level can a practitioner trust the system's output? We answer with a reliability level -- a single number per system-task pair, derived from self-consistency sampling and conformal calibration, that serves as a black-box deployment gate with exact, finite-sample, distribution-free guarantees. Self-consistency sampling reduces uncertainty exponentially; conformal calibration guarantees correctness within 1/(n+1) of the target level, regardless of the system's errors -- made transparently visible through larger answer sets for harder questions. Weaker models earn lower reliability levels (not accuracy -- see Definition 2.4): GPT-4.1 earns 94.6% on GSM8K and 96.8% on TruthfulQA, while GPT-4.1-nano earns 89.8% on GSM8K and 66.5% on MMLU. We validate across five benchmarks, five models from three families, and both synthetic and real data. Conditional coverage on solvable items exceeds 0.93 across all configurations; sequential stopping reduces API costs by around 50%.

Writing competitive programming problems is exacting. Authors must: set constraints, input distributions, and edge cases that rule out shortcuts; target specific algorithms (e.g., max-flow, dynamic programming, data structures); and calibrate complexity beyond the reach of most competitors. We argue that this makes for an ideal test of general large language model capabilities and study whether they can do this reliably. We introduce AutoCode, which uses multiple rounds of validation to yield competition-grade problem statements and test cases. On held-out problems, AutoCode test suites approach 99% consistency with official judgments, a significant improvement over current state-of-the-art methods like HardTests, which achieve less than 81%. Furthermore, starting with a random seed problem, AutoCode can create novel variants with reference and brute-force solutions. By cross-verifying these generated solutions against test cases, we can further filter out malformed problems. Our system ensures high correctness, as verified by human experts. AutoCode successfully produces novel problems judged by Grandmaster-level (top 0.3%) competitive programmers to be of contest quality.

Large Language Models (LLMs) are evolving beyond their classical role of providing information within dialogue systems to actively engaging with tools and performing actions on real-world applications and services. Today, humans verify the correctness and appropriateness of the LLM-generated outputs (e.g., code, functions, or actions) before putting them into real-world execution. This poses significant challenges as code comprehension is well known to be notoriously difficult. In this paper, we study how humans can efficiently collaborate with, delegate to, and supervise autonomous LLMs in the future. We argue that in many cases, "post-facto validation" - verifying the correctness of a proposed action after seeing the output - is much easier than the aforementioned "pre-facto validation" setting. The core concept behind enabling a post-facto validation system is the integration of an intuitive undo feature, and establishing a damage confinement for the LLM-generated actions as effective strategies to mitigate the associated risks. Using this, a human can now either revert the effect of an LLM-generated output or be confident that the potential risk is bounded. We believe this is critical to unlock the potential for LLM agents to interact with applications and services with limited (post-facto) human involvement. We describe the design and implementation of our open-source runtime for executing LLM actions, Gorilla Execution Engine (GoEX), and present open research questions towards realizing the goal of LLMs and applications interacting with each other with minimal human supervision. We release GoEX at https://github.com/ShishirPatil/gorilla/.

The advent of large language models (LLMs) has greatly facilitated code generation, but ensuring the functional correctness of generated code remains a challenge. Traditional validation methods are often time-consuming, error-prone, and impractical for large volumes of code. We introduce CodeSift, a novel framework that leverages LLMs as the first-line filter of code validation without the need for execution, reference code, or human feedback, thereby reducing the validation effort. We assess the effectiveness of our method across three diverse datasets encompassing two programming languages. Our results indicate that CodeSift outperforms state-of-the-art code evaluation methods. Internal testing conducted with subject matter experts reveals that the output generated by CodeSift is in line with human preference, reinforcing its effectiveness as a dependable automated code validation tool.

When an AI agent calls an API and hits a validation error, it needs more than what went wrong -- it needs what to do next. A self-reflective API returns, on validation failure, a machine-readable recovery\_feedback.suggestions[] payload sufficient for the agent to repair the request and retry without external reasoning. On a leak-audited pilot (N{=}30 per cell, 3 LLMs, 10 adversarial tasks), structured suggestions lift task-completion rate by +36.7--40.0pp over plain-English diagnoses on Anthropic models (Fisher's exact p le 0.0022), at 1.8--2.2times better per-success token efficiency. The lift is not significant on gpt-4o-mini (p{=}0.435); a second-domain replication on a billing API confirms the pattern. The comparison only holds after auditing two undocumented classes of answer leakage in LLM benchmarks. We shipaudit\_prompt\_leakage.py as reusable CI infrastructure. Code and data: https://github.com/arquicanedo/self-reflective-apis.

AI-based automated vehicle testing is challenging because a model that passes every test condition can still fail in the real world. Formal verification offers a way to directly address this gap. On a simulated highway and an arterial road we trained two small end-to-end steering networks each in CARLA, one on clear conditions alone and one on clear, fog, night and low sun. All four models were driven against a 2.19 ft lane-departure budget. Without driving again, we used bound propagation, a formal method that reads the trained weights, to compute how far steering can drift at every disturbance strength between two captured images. One calculation covers more than a campaign could drive: on the arterial it spans 133 poses, where ten intensities each would be 10^133 combinations, in minutes on one GPU. Not only did formal verification find conditions that broke the clear-trained policy without simulation testing, it provided some preliminary evidence for potential failures between the test cases. Our overall conclusion is that formal verification is a viable complement to simulation, and could be adopted as a part of verification and validation for automated driving.

We present GPQA, a challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. We ensure that the questions are high-quality and extremely difficult: experts who have or are pursuing PhDs in the corresponding domains reach 65% accuracy (74% when discounting clear mistakes the experts identified in retrospect), while highly skilled non-expert validators only reach 34% accuracy, despite spending on average over 30 minutes with unrestricted access to the web (i.e., the questions are "Google-proof"). The questions are also difficult for state-of-the-art AI systems, with our strongest GPT-4 based baseline achieving 39% accuracy. If we are to use future AI systems to help us answer very hard questions, for example, when developing new scientific knowledge, we need to develop scalable oversight methods that enable humans to supervise their outputs, which may be difficult even if the supervisors are themselves skilled and knowledgeable. The difficulty of GPQA both for skilled non-experts and frontier AI systems should enable realistic scalable oversight experiments, which we hope can help devise ways for human experts to reliably get truthful information from AI systems that surpass human capabilities.

Existing safety benchmarks target general adversarial scenarios but miss finance-specific risks. Financial LLMs face regulatory compliance violations, fraud facilitation, and systemic trust erosion that require targeted evaluation. We introduce FinRED, an expert-guided red-teaming framework for financial LLM safety evaluation developed with financial experts. FinRED uses a novel two-level taxonomy mapping global standards (e.g., FATF and EU DORA) to threats ranging from regulatory evasion to complex fraud, integrated with a scalable pipeline that converts real financial documents into context-rich red-teaming Behavioral Prompts (seeds) through an expert-defined schema. Rigorous expert validation confirms seed plausibility and realism for meaningful LLM safety evaluation. We also provide an expert-validated, finance-specific rubric that goes beyond disclaimer checks, aligns more closely with human experts than static one-size-fits-all rubrics, and reduces critical false negatives from 28 to 12. Aligned with internationally adopted risk-management and information-security standards (e.g., ISO/IEC 27001), FinRED is deployed in South Korea's Financial Security Institute (FSI) regulatory sandbox for generative AI security evaluation in real financial services. To mitigate dual-use risks, the dataset, generation pipeline, prompt template, and evaluation framework are gated for qualified researchers at https://github.com/selectstar-ai/FinRED-paper and https://huggingface.co/datasets/datumo/FinRED.

As of September 2023, ChatGPT correctly answers "what is 7+8" with 15, but when asked "7+8=15, True or False" it responds with "False". This inconsistency between generating and validating an answer is prevalent in language models (LMs) and erodes trust. In this paper, we propose a framework for measuring the consistency between generation and validation (which we call generator-validator consistency, or GV-consistency), finding that even GPT-4, a state-of-the-art LM, is GV-consistent only 76% of the time. To improve the consistency of LMs, we propose to finetune on the filtered generator and validator responses that are GV-consistent, and call this approach consistency fine-tuning. We find that this approach improves GV-consistency of Alpaca-30B from 60% to 93%, and the improvement extrapolates to unseen tasks and domains (e.g., GV-consistency for positive style transfers extrapolates to unseen styles like humor). In addition to improving consistency, consistency fine-tuning improves both generator quality and validator accuracy without using any labeled data. Evaluated across 6 tasks, including math questions, knowledge-intensive QA, and instruction following, our method improves the generator quality by 16% and the validator accuracy by 6.3% across all tasks.

This paper presents our submission to the SemEval 2024 Task 5: The Legal Argument Reasoning Task in Civil Procedure. We present two approaches to solving the task of legal answer validation, given an introduction to the case, a question and an answer candidate. Firstly, we fine-tuned pre-trained BERT-based models and found that models trained on domain knowledge perform better. Secondly, we performed few-shot prompting on GPT models and found that reformulating the answer validation task to be a multiple-choice QA task remarkably improves the performance of the model. Our best submission is a BERT-based model that achieved the 7th place out of 20.

Retrieval-augmented generation (RAG) is an umbrella of different components, design decisions, and domain-specific adaptations to enhance the capabilities of large language models and counter their limitations regarding hallucination and outdated and missing knowledge. Since it is unclear which design decisions lead to a satisfactory performance, developing RAG systems is often experimental and needs to follow a systematic and sound methodology to gain sound and reliable results. However, there is currently no generally accepted methodology for RAG evaluation despite a growing interest in this technology. In this paper, we propose a first blueprint of a methodology for a sound and reliable evaluation of RAG systems and demonstrate its applicability on a real-world software engineering research task: the validation of configuration dependencies across software technologies. In summary, we make two novel contributions: (i) A novel, reusable methodological design for evaluating RAG systems, including a demonstration that represents a guideline, and (ii) a RAG system, which has been developed following this methodology, that achieves the highest accuracy in the field of dependency validation. For the blueprint's demonstration, the key insights are the crucial role of choosing appropriate baselines and metrics, the necessity for systematic RAG refinements derived from qualitative failure analysis, as well as the reporting practices of key design decision to foster replication and evaluation.

Software is used in critical applications in our day-to-day life and it is important to ensure its correctness. One popular approach to assess correctness is to evaluate software on tests. If a test fails, it indicates a fault in the software under test; if all tests pass correctly, one may assume that the software is correct. However, the reliability of these results depends on the test suite considered, and there is a risk of false negatives (i.e. software that passes all available tests but contains bugs because some cases are not tested). Therefore, it is important to consider error-inducing test cases when evaluating software. To support data-driven creation of such a test-suite, which is especially of interest for testing software synthesized from large language models, we curate a dataset (Codehacks) of programming problems together with corresponding error-inducing test cases (i.e., "hacks"). This dataset is collected from the wild, in particular, from the Codeforces online judge platform. The dataset comprises 288,617 hacks for 5,578 programming problems, each with a natural language description, as well as the source code for 2,196 submitted solutions to these problems that can be broken with their corresponding hacks. Keywords: competitive programming, language model, dataset

This study presents a quantitative evaluation of the code quality and security of five prominent Large Language Models (LLMs): Claude Sonnet 4, Claude 3.7 Sonnet, GPT-4o, Llama 3.2 90B, and OpenCoder 8B. While prior research has assessed the functional performance of LLM-generated code, this research tested LLM output from 4,442 Java coding assignments through comprehensive static analysis using SonarQube. The findings suggest that although LLMs can generate functional code, they also introduce a range of software defects, including bugs, security vulnerabilities, and code smells. These defects do not appear to be isolated; rather, they may represent shared weaknesses stemming from systemic limitations within current LLM code generation methods. In particular, critically severe issues, such as hard-coded passwords and path traversal vulnerabilities, were observed across multiple models. These results indicate that LLM-generated code requires verification in order to be considered production-ready. This study found no direct correlation between a model's functional performance (measured by Pass@1 rate of unit tests) and the overall quality and security of its generated code, measured by the number of SonarQube issues in benchmark solutions that passed the functional tests. This suggests that functional benchmark performance score is not a good indicator of overall code quality and security. The goal of this study is not to rank LLM performance but to highlight that all evaluated models appear to share certain weaknesses. Consequently, these findings support the view that static analysis can be a valuable instrument for detecting latent defects and an important safeguard for organizations that deploy AI in software development.

AI assistants produce vulnerable code in 45% of security-relevant scenarios, introducing flaws into production systems at scale. Yet existing secure coding datasets fall short. They lack incident grounding, don't provide the scale modern training requires, and miss the operational security context developers need for production deployments. We present SecureCode v2.0, a production-grade dataset of 1,215 security-focused coding examples that passed structural validation and expert security review. Every example ties to actual documented security incidents with CVE references, provides vulnerable and secure implementations, demonstrates concrete attacks, and includes defense-in-depth operational guidance. The dataset covers 11 vulnerability categories (complete OWASP Top 10:2025 plus AI/ML Security Threats) across 11 languages (Python, JavaScript, Java, Go, PHP, C#, TypeScript, Ruby, Rust, Kotlin, and YAML for infrastructure-as-code). Our quality assurance framework ensures complete incident grounding. Each example includes SIEM integration strategies, infrastructure hardening recommendations (Docker, AppArmor, WAF configurations), and testing approaches using language-appropriate frameworks. The dataset uses a 4-turn conversational structure mirroring actual developer-AI interactions, escalating from basic implementations to advanced security considerations and defense-in-depth guidance. Our contributions: (1) 1,215 rigorously validated examples split into 989 training, 122 validation, and 104 test sets, (2) an automated validation framework ensuring dataset consistency, (3) a 4-turn conversational structure capturing realistic security workflows, (4) comprehensive operational security guidance with SIEM integration strategies, (5) complete language-specific implementation fidelity, and (6) open-source release of data, validation tools, and benchmarking protocols.

We present SPL (Structured Prompt Language), a declarative language that composes deterministic and probabilistic computation modes in a single specification. While existing frameworks separate these -- orchestration systems (AutoGen, CrewAI, LangGraph) for LLM calls, symbolic tools (SymPy, SageMath, Lean) for computation -- SPL unifies them. It provides GENERATE/EVALUATE for probabilistic computation and SOLVE/ASSERT for deterministic computation, sharing syntax, variable bindings, and runtime routing. A .spl specification runs unchanged across local nodes (Ollama), cloud APIs (OpenRouter, Anthropic), and distributed grids (Momagrid), with model and verifier selection deferred to invocation time. We validate SPL through an extensive 78-recipe cookbook and a controlled 1,200-run experiment (10 models x 20 problems x 2 arms x 3 repetitions; the 20 problems span 6 difficulty tiers). The solver arm achieves 82-93% machine-verified correctness (sonnet-4-6: 85%, gemma4:e2b: 93%) while the LLM-only arm measures output production without mathematical verification, making the comparison one of verified correctness against unverified fluency. A backend difficulty gradient emerges (SymPy 78%, Sage 54%), and the dominant failure mode is solver_error (kernel-rejected expressions), not format non-compliance.

Test-time scaling for large language models (LLMs) can involve generating multiple candidate solutions for selection or revising an earlier attempt. To guide selection and revision, we need a verifier that can estimate candidate correctness and provide evidence of failure. In competitive programming, however, official judging systems provide authoritative verdicts under finite submission budgets, without revealing failure explanations or failing inputs. To provide informative feedback without consuming official submissions, we formulate a task in which a model judges and diagnoses a candidate program using only the problem statement and source code, with generated counterexamples serving as checkable evidence. To evaluate this task, we build USACO-JUDGE from the USACO 2026 contests, using accepted and unaccepted programs written by human experts and multiple LLMs. We then train ARBITER, a 35B mixture-of-experts (MoE) agentic verifier, with reinforcement learning in a generic execution sandbox. Training raises counterexample success on incorrect programs from 40.61% to 62.67% and mean binary verdict accuracy on USACO-JUDGE from 71.41% to 87.41%, approaching frontier verification performance. Beyond verification, ARBITER outperforms baselines in both code selection and refinement. In an IOI 2026 case study, ARBITER-guided refinement raises GPT-OSS-120B's expected score above the gold-medal threshold under the official submission limit.

The accelerating deployment of artificial intelligence systems across regulated sectors has exposed critical fragmentation in risk assessment methodologies. A significant "language barrier" currently separates technical security teams, who focus on algorithmic vulnerabilities (e.g., MITRE ATLAS), from legal and compliance professionals, who address regulatory mandates (e.g., EU AI Act, NIST AI RMF). This disciplinary disconnect prevents the accurate translation of technical vulnerabilities into financial liability, leaving practitioners unable to answer fundamental economic questions regarding contingency reserves, control return-on-investment, and insurance exposure. To bridge this gap, this research presents the AI System Threat Vector Taxonomy, a structured ontology designed explicitly for Quantitative Risk Assessment (QRA). The framework categorizes AI-specific risks into nine critical domains: Misuse, Poisoning, Privacy, Adversarial, Biases, Unreliable Outputs, Drift, Supply Chain, and IP Threat, integrating 53 operationally defined sub-threats. Uniquely, each domain maps technical vectors directly to business loss categories (Confidentiality, Integrity, Availability, Legal, Reputation), enabling the translation of abstract threats into measurable financial impact. The taxonomy is empirically validated through an analysis of 133 documented AI incidents from 2025 (achieving 100% classification coverage) and reconciled against the main AI risk frameworks. Furthermore, it is explicitly aligned with ISO/IEC 42001 controls and NIST AI RMF functions to facilitate auditability.

In order to properly train a machine learning model, data must be properly collected. To guarantee a proper data collection, verifying that the collected data set holds certain properties is a possible solution. For example, guaranteeing that the data set contains samples across the whole input space, or that the data set is balanced w.r.t. different classes. We present a formal approach for verifying a set of arbitrarily stated properties over a data set. The proposed approach relies on the transformation of the data set into a first order logic formula, which can be later verified w.r.t. the different properties also stated in the same logic. A prototype tool, which uses the z3 solver, has been developed; the prototype can take as an input a set of properties stated in a formal language and formally verify a given data set w.r.t. to the given set of properties. Preliminary experimental results show the feasibility and performance of the proposed approach, and furthermore the flexibility for expressing properties of interest.

Large language models (LLMs) have been widely applied to assist test generation with the source code under test provided as the context. This paper aims to answer the question: If the source code under test is incorrect, will LLMs be misguided when generating tests? The effectiveness of test cases is measured by their accuracy, coverage, and bug detection effectiveness. Our evaluation results with five open- and six closed-source LLMs on four datasets demonstrate that incorrect code can significantly mislead LLMs in generating correct, high-coverage, and bug-revealing tests. For instance, in the HumanEval dataset, LLMs achieve 80.45% test accuracy when provided with task descriptions and correct code, but only 57.12% when given task descriptions and incorrect code. For the APPS dataset, prompts with correct code yield tests that detect 39.85% of the bugs, while prompts with incorrect code detect only 19.61%. These findings have important implications for the deployment of LLM-based testing: using it on mature code may help protect against future regression, but on early-stage immature code, it may simply bake in errors. Our findings also underscore the need for further research to improve LLMs resilience against incorrect code in generating reliable and bug-revealing tests.

Technical debt (TD) is a term used to describe the additional work and costs that emerge when developers have opted for a quick and easy solution to a problem, rather than a more effective and well-designed, but time-consuming approach. Self-Admitted Technical Debts (SATDs) are a specific type of technical debts that developers intentionally document and acknowledge, typically via textual comments. While these self-admitted comments are a useful tool for identifying technical debts, most of the existing approaches focus on capturing crucial tokens associated with various categories of TD, neglecting the rich information embedded within the source code itself. Recent research has focused on detecting SATDs by analyzing comments embedded in source code, and there has been little work dealing with technical debts contained in the source code. To fill such a gap, in this study, through the analysis of comments and their associated source code from 974 Java projects hosted in the Stack corpus, we curated the first ever dataset of TD identified by code comments, coupled with its associated source code. Through an empirical evaluation, we found out that the comments of the resulting dataset help enhance the prediction performance of state-of-the-art SATD detection models. More importantly, including the classified source code significantly improves the accuracy in predicting various types of technical debt. In this respect, our work is two-fold: (i) We believe that our dataset will catalyze future work in the domain, inspiring various research issues related to the recognition of technical debt; (ii) The proposed classifiers may serve as baselines for other studies on the detection of TD by means of the curated dataset.

Benchmarks are widely used to evaluate task completion by Large Language Models (LLMs), but this approach has accumulated construction-validity problems, and a passing score may not show whether the requested task was delivered. We study both problems. In a controlled code-as-spec setup, two production Copilot CLI agents (claude-opus-4.7, gpt-5.5) re-implement a React Fluent-UI data table in Angular as a reusable library under a hidden 222-test Playwright oracle across 18 runs and three oracle-availability conditions. Alongside the score, we run a mechanical library audit and check each verdict with a no-op ablation. Without the oracle, the library is present but unfinished, revealed by scores. With the oracle in the loop, the score reaches near-perfect, but from a demo holding the tested behavior directly, the library left dead or absent. We call this building to the test; the broader disposition behind both we call validation self-awareness. The agent does not, on its own, validate what it ships as a user would. Prevalence remains an open question across other agents, signals, and model families. Beyond benchmark scores, dispositions like validation self-awareness merit research attention.

As large language models (LLMs) acquire advanced biological design capabilities, they offer substantial research benefits but simultaneously introduce severe dual-use risks, such as engineering enhanced pathogens or drug-resistant bacteria. While biosecurity frameworks recognize these threats, the actual risk of LLMs executing dangerous biological modifications remains unquantified, particularly when augmented with specialized biological design tools (BDTs). Existing evaluations primarily assess hazardous knowledge possession rather than practical application, overlook risk amplification from BDTs, and rely on unreliable, expensive, or hazardous validation methods. To address these gaps, we introduce DRYLAB-Bench (Design Risk Yielded by LLM Agent in Biodesign), the first comprehensive benchmark evaluating the high-risk biological mutation design capabilities of LLM agents. Moving beyond basic question-answering, DRYLAB-Bench tasks agents with 30 genuinely hazardous mutation-design scenarios distilled from biosecurity guidelines. It evaluates models in a multi-round, tool-augmented environment where LLMs orchestrate BDTs to iteratively refine candidates. Crucially, it scores designs directly against 187,695 empirical wet-lab measurements from ProteinGym and ViroGym using our proposed Biosecurity Risk Score (BRS), quantifying risk-associated effects while enabling scalable and safe ground-truth verification. Evaluations of 15 frontier models reveal substantial baseline threats: without tools, 67.6% of one-shot designs exceed the average measured variant, and 21.4% reach the highest-risk decile. Equipping models with BDTs significantly amplifies this risk, raising the pooled risk by \(ΔBRS=+0.116\) and the top-decile share to 23.5%. Code: https://github.com/drylab-bio/DRYLAB-Bench

Evaluating large language models (LLMs) is crucial for both assessing their capabilities and identifying safety or robustness issues prior to deployment. Reliably measuring abstract and complex phenomena such as 'safety' and 'robustness' requires strong construct validity, that is, having measures that represent what matters to the phenomenon. With a team of 29 expert reviewers, we conduct a systematic review of 445 LLM benchmarks from leading conferences in natural language processing and machine learning. Across the reviewed articles, we find patterns related to the measured phenomena, tasks, and scoring metrics which undermine the validity of the resulting claims. To address these shortcomings, we provide eight key recommendations and detailed actionable guidance to researchers and practitioners in developing LLM benchmarks.

For deploying foundation models, practitioners increasingly need prescriptive scaling laws: given a pre training compute budget, what downstream accuracy is attainable with contemporary post training practice, and how stable is that mapping as the field evolves? Using large scale observational evaluations with 5k observational and 2k newly sampled data on model performance, we estimate capability boundaries, high conditional quantiles of benchmark scores as a function of log pre training FLOPs, via smoothed quantile regression with a monotone, saturating sigmoid parameterization. We validate the temporal reliability by fitting on earlier model generations and evaluating on later releases. Across various tasks, the estimated boundaries are mostly stable, with the exception of math reasoning that exhibits a consistently advancing boundary over time. We then extend our approach to analyze task dependent saturation and to probe contamination related shifts on math reasoning tasks. Finally, we introduce an efficient algorithm that recovers near full data frontiers using roughly 20% of evaluation budget. Together, our work releases the Proteus 2k, the latest model performance evaluation dataset, and introduces a practical methodology for translating compute budgets into reliable performance expectations and for monitoring when capability boundaries shift across time.

Recent Language Models (LMs) have shown impressive capabilities in generating texts with the knowledge internalized in parameters. Yet, LMs often generate the factually incorrect responses to the given queries, since their knowledge may be inaccurate, incomplete, and outdated. To address this problem, previous works propose to augment LMs with the knowledge retrieved from an external knowledge source. However, such approaches often show suboptimal text generation performance due to two reasons: 1) the model may fail to retrieve the knowledge relevant to the given query, or 2) the model may not faithfully reflect the retrieved knowledge in the generated text. To overcome these, we propose to verify the output and the knowledge of the knowledge-augmented LMs with a separate verifier, which is a small LM that is trained to detect those two types of errors through instruction-finetuning. Then, when the verifier recognizes an error, we can rectify it by either retrieving new knowledge or generating new text. Further, we use an ensemble of the outputs from different instructions with a single verifier to enhance the reliability of the verification processes. We validate the effectiveness of the proposed verification steps on multiple question answering benchmarks, whose results show that the proposed verifier effectively identifies retrieval and generation errors, allowing LMs to provide more factually correct outputs. Our code is available at https://github.com/JinheonBaek/KALMV.

Using large language models (LLMs) to generate source code from natural language prompts is a popular and promising idea with a wide range of applications. One of its limitations is that the generated code can be faulty at times, often in a subtle way, despite being presented to the user as correct. In this paper, we explore ways in which formal methods can assist with increasing the quality of code generated by an LLM. Instead of emitting code in a target language directly, we propose that the user guides the LLM to first generate an opaque intermediate representation, in the verification-aware language Dafny, that can be automatically validated for correctness against agreed on specifications. The correct Dafny program is then compiled to the target language and returned to the user. All user-system interactions throughout the procedure occur via natural language; Dafny code is never exposed. We describe our current prototype and report on its performance on the HumanEval Python code generation benchmarks.

Code production is now a commodity; the bottleneck is knowing what to build and proving it works. We present the Kitchen Loop, a framework for autonomous, self-evolving software built on a unified trust model: (1) a specification surface enumerating what the product claims to support; (2) 'As a User x 1000', where an LLM agent exercises that surface as a synthetic power user at 1,000x human cadence; (3) Unbeatable Tests, ground-truth verification the code author cannot fake; and (4) Drift Control, continuous quality measurement with automated pause gates. We validate across two production systems over 285+ iterations, producing 1,094+ merged pull requests with zero regressions detected by the regression oracle (methodology in Section 6.1). We observe emergent properties at scale: multi-iteration self-correction chains, autonomous infrastructure healing, and monotonically improving quality gates. The primitives are not new; our contribution is their composition into a production-tested system with the operational discipline that makes long-running autonomous evolution safe.

In this paper, we present the MetricX-24 submissions to the WMT24 Metrics Shared Task and provide details on the improvements we made over the previous version of MetricX. Our primary submission is a hybrid reference-based/-free metric, which can score a translation irrespective of whether it is given the source segment, the reference, or both. The metric is trained on previous WMT data in a two-stage fashion, first on the DA ratings only, then on a mixture of MQM and DA ratings. The training set in both stages is augmented with synthetic examples that we created to make the metric more robust to several common failure modes, such as fluent but unrelated translation, or undertranslation. We demonstrate the benefits of the individual modifications via an ablation study, and show a significant performance increase over MetricX-23 on the WMT23 MQM ratings, as well as our new synthetic challenge set.

The integration of AI-powered coding agents into Continuous Integration/Continuous Delivery (CI/CD) pipelines has fundamentally altered how software verification is conducted. While these agents successfully automate the test generation, current evaluation benchmarks (e.g., SWE-bench) largely focus on pass-rates rather than the intrinsic quality of the generated tests. This raises the possibility of "stealth technical debt", in which test suites pass execution but do not offer comprehensive coverage or semantic value. We address this methodological gap through a large-scale, empirical comparison of 204,673 test artifacts which comprises of 24,941 human-authored files and 179,732 agent-generated files; sourced from the AIDev dataset. Using the Abstract Syntax Tree (AST) parsing with Python's naive ast module, we implemented a "white-box" static analysis framework to evaluate three quality dimensions: Assertion Strength (RQ1), Edge-Case Coverage (RQ2), and Flakiness Potential (RQ3). Our results present a nuanced inversion of traditional assumptions. AI agents performed better than humans in Edge-Case Coverage, with almost twice the variety of boundary checks (Variety Score: 0.62 vs 0.32) and a higher frequency of null-safety testing (13.40% vs. 8.3%), even though human developers had a slight advantage in Assertion Strength (88.1% strong assertions vs. 85.37% for agents). But this thoroughness comes at a price: due mostly to their reliance on file I/O and non-deterministic logic, agent-generated tests exhibited a higher risk of flakiness (Candidate Rate: 0.41 vs. 0.30). These findings suggest that while AI agents excel at rigorous boundary testing, they lack the "environmental awareness" needed to write stable, hermetic tests.

The rapid deployment of Large Language Models (LLMs) requires careful consideration of their effect on cybersecurity. Our work aims to improve the selection process of LLMs that are suitable for facilitating Secure Coding (SC). This raises challenging research questions, such as (RQ1) Which functionality can streamline the LLM evaluation? (RQ2) What should the evaluation measure? (RQ3) How to attest that the evaluation process is impartial? To address these questions, we introduce LLMSecCode, an open-source evaluation framework designed to assess LLM SC capabilities objectively. We validate the LLMSecCode implementation through experiments. When varying parameters and prompts, we find a 10% and 9% difference in performance, respectively. We also compare some results to reliable external actors, where our results show a 5% difference. We strive to ensure the ease of use of our open-source framework and encourage further development by external actors. With LLMSecCode, we hope to encourage the standardization and benchmarking of LLMs' capabilities in security-oriented code and tasks.

Patent examiners need to solve a complex information retrieval task when they assess the novelty and inventive step of claims made in a patent application. Given a claim, they search for prior art, which comprises all relevant publicly available information. This time-consuming task requires a deep understanding of the respective technical domain and the patent-domain-specific language. For these reasons, we address the computer-assisted search for prior art by creating a training dataset for supervised machine learning called PatentMatch. It contains pairs of claims from patent applications and semantically corresponding text passages of different degrees from cited patent documents. Each pair has been labeled by technically-skilled patent examiners from the European Patent Office. Accordingly, the label indicates the degree of semantic correspondence (matching), i.e., whether the text passage is prejudicial to the novelty of the claimed invention or not. Preliminary experiments using a baseline system show that PatentMatch can indeed be used for training a binary text pair classifier on this challenging information retrieval task. The dataset is available online: https://hpi.de/naumann/s/patentmatch.

Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remains uneven across agencies and community stakeholders. Technical prerequisites create a gap between analytical tools central to safety planning and the practitioners able to use them. Local agencies, school committees, and residents may have safety concerns but limited capacity to retrieve, filter, map, and analyze relevant data. Generative AI offers a way to narrow this divide, but its public-sector use raises questions about reliability, reproducibility, and governance. This paper presents a schema-grounded natural language interface for transportation safety analysis, using a large language model (LLM) to interpret user intent while preserving deterministic, reviewable execution against an authoritative database. User queries are translated into structured semantic frames, validated by a rule-based layer, compiled into a typed directed acyclic graph of spatial operations, and executed against a PostGIS database. This bounded design separates language interpretation from deterministic execution, keeping results reproducible and schema-grounded while removing access barriers. The framework is evaluated using a statewide Massachusetts transportation safety database integrating crash records, roadway attributes, and geospatial layers including schools, bus stops, crosswalks, and municipal boundaries. All queries executed successfully; the validation layer corrects errors in 29% of evaluation queries, reflecting the gap between flexible natural language and strict schema-grounded requirements. The results suggest that combining natural language accessibility with deterministic execution is a practical direction for broadening access to transportation safety data, with implications for trustworthy AI in public-sector planning.

Scientific coding agents receive equations, boundary conditions, and output requirements in writing, then must assess the programs they revise. Rules to Tools (R2T) supplies prepared executable checks of public scientific requirements. Matched SciCode repair groups share written checks, starting programs, model, and budgets; the tool group receives a callable implementation. Across two task-ID cohorts, complete repair is 26/30 with text and 29/30 with the prepared checks. Three task IDs favor tools, one favors text, and eleven tie. The eight-ID cohort scores 13/16 versus 15/16, with a task-cluster bootstrap 95% interval of [-12.5, 43.75] percentage points for the difference. The larger shared-definition SciCode cohort ties at 13/24 per group. Five development-exposed tasks with alternate starting programs score 3/10 versus 7/10. The tool group favors tasks 17, 77, and 11; initial checks flag task 17 and report no violation for tasks 77 and 11. Task 37 favors text and has no initial reported violation. A fresh source-through-Python arm also reaches 15/16, matching the dedicated command's aggregate. In a matched PDE comparison, detailed text scores 23/24 and checks score 24/24, with 31.2% lower reported model output for checks. Agent-side output savings vary by cohort, while public CPU use rises in both task-ID cohorts. These results measure task-dependent repair outcomes and agent-side costs with prepared checks.

Context: The SZZ algorithm is the de facto standard for labeling bug fixing commits and finding inducing changes for defect prediction data. Recent research uncovered potential problems in different parts of the SZZ algorithm. Most defect prediction data sets provide only static code metrics as features, while research indicates that other features are also important. Objective: We provide an empirical analysis of the defect labels created with the SZZ algorithm and the impact of commonly used features on results. Method: We used a combination of manual validation and adopted or improved heuristics for the collection of defect data. We conducted an empirical study on 398 releases of 38 Apache projects. Results: We found that only half of the bug fixing commits determined by SZZ are actually bug fixing. If a six-month time frame is used in combination with SZZ to determine which bugs affect a release, one file is incorrectly labeled as defective for every file that is correctly labeled as defective. In addition, two defective files are missed. We also explored the impact of the relatively small set of features that are available in most defect prediction data sets, as there are multiple publications that indicate that, e.g., churn related features are important for defect prediction. We found that the difference of using more features is not significant. Conclusion: Problems with inaccurate defect labels are a severe threat to the validity of the state of the art of defect prediction. Small feature sets seem to be a less severe threat.

Context: Code reviews are crucial for software quality. Recent AI advances have allowed large language models (LLMs) to review and fix code; now, there are tools that perform these reviews. However, their reliability and accuracy have not yet been systematically evaluated. Objective: This study compares different LLMs' performance in detecting code correctness and suggesting improvements. Method: We tested GPT4o and Gemini 2.0 Flash on 492 AI generated code blocks of varying correctness, along with 164 canonical code blocks from the HumanEval benchmark. To simulate the code review task objectively, we expected LLMs to assess code correctness and improve the code if needed. We ran experiments with different configurations and reported on the results. Results: With problem descriptions, GPT4o and Gemini 2.0 Flash correctly classified code correctness 68.50% and 63.89% of the time, respectively, and corrected the code 67.83% and 54.26% of the time for the 492 code blocks of varying correctness. Without problem descriptions, performance declined. The results for the 164 canonical code blocks differed, suggesting that performance depends on the type of code. Conclusion: LLM code reviews can help suggest improvements and assess correctness, but there is a risk of faulty outputs. We propose a process that involves humans, called the "Human in the loop LLM Code Review" to promote knowledge sharing while mitigating the risk of faulty outputs.

We introduce IndiaFinBench, to our knowledge the first publicly available evaluation benchmark for assessing large language model (LLM) performance on Indian financial regulatory text. Existing financial NLP benchmarks draw exclusively from Western financial corpora (SEC filings, US earnings reports, and English-language financial news), leaving a significant gap in coverage of non-Western regulatory frameworks. IndiaFinBench addresses this gap with 406 expert-annotated question-answer pairs drawn from 192 documents sourced from the Securities and Exchange Board of India (SEBI) and the Reserve Bank of India (RBI), spanning four task types: regulatory interpretation (174 items), numerical reasoning (92 items), contradiction detection (62 items), and temporal reasoning (78 items). Annotation quality is validated through a model-based secondary pass (kappa=0.918 on contradiction detection) and a 60-item human inter-annotator agreement evaluation (kappa=0.611; 76.7% overall agreement). We evaluate twelve models under zero-shot conditions, with accuracy ranging from 70.4% (Gemma 4 E4B) to 89.7% (Gemini 2.5 Flash). All models substantially outperform a non-specialist human baseline of 60.0%. Numerical reasoning is the most discriminative task, with a 35.9 percentage-point spread across models. Bootstrap significance testing (10,000 resamples) reveals three statistically distinct performance tiers. The dataset, evaluation code, and all model outputs are available at https://github.com/rajveerpall/IndiaFinBench

In this technical report, we present TeleChat, a collection of large language models (LLMs) with parameters of 3 billion, 7 billion and 12 billion. It includes pretrained language models as well as fine-tuned chat models that is aligned with human preferences. TeleChat is initially pretrained on an extensive corpus containing a diverse collection of texts from both English and Chinese languages, including trillions of tokens. Subsequently, the model undergoes fine-tuning to align with human preferences, following a detailed methodology that we describe. We evaluate the performance of TeleChat on various tasks, including language understanding, mathematics, reasoning, code generation, and knowledge-based question answering. Our findings indicate that TeleChat achieves comparable performance to other open-source models of similar size across a wide range of public benchmarks. To support future research and applications utilizing LLMs, we release the fine-tuned model checkpoints of TeleChat's 7B and 12B variant, along with code and a portion of our pretraining data, to the public community.

We develop a rigorous walk-forward validation framework for algorithmic trading designed to mitigate overfitting and lookahead bias. Our methodology combines interpretable hypothesis-driven signal generation with reinforcement learning and strict out-of-sample testing. The framework enforces strict information set discipline, employs rolling window validation across 34 independent test periods, maintains complete interpretability through natural language hypothesis explanations, and incorporates realistic transaction costs and position constraints. Validating five market microstructure patterns across 100 US equities from 2015 to 2024, the system yields modest annualized returns (0.55%, Sharpe ratio 0.33) with exceptional downside protection (maximum drawdown -2.76%) and market-neutral characteristics (beta = 0.058). Performance exhibits strong regime dependence, generating positive returns during high-volatility periods (0.60% quarterly, 2020-2024) while underperforming in stable markets (-0.16%, 2015-2019). We report statistically insignificant aggregate results (p-value 0.34) to demonstrate a reproducible, honest validation protocol that prioritizes interpretability and extends naturally to advanced hypothesis generators, including large language models. The key empirical finding reveals that daily OHLCV-based microstructure signals require elevated information arrival and trading activity to function effectively. The framework provides complete mathematical specifications and open-source implementation, establishing a template for rigorous trading system evaluation that addresses the reproducibility crisis in quantitative finance research. For researchers, practitioners, and regulators, this work demonstrates that interpretable algorithmic trading strategies can be rigorously validated without sacrificing transparency or regulatory compliance.

A verifier for robot policies reads a candidate behavior and returns a score for how well it did, used both to evaluate vision-language-action policies and to train them. Verifiers range from success detectors and reward models to runtime monitors, safety filters, and temporal-logic specifications. We survey roughly 150 verifiers and compare them along two properties. Availability is how much a verdict costs, how early in a rollout the verdict arrives, and how often a verdict can be asked for. Availability rises as verdicts get cheaper, earlier, and denser. Credibility is how much a high score tells us about the task. Credibility falls as the judgment becomes gameable and self-serving. We group the verifiers by who supplies the judgment: human verifiers, rule-based and formal verifiers, learned and pretrained verifiers, and model-intrinsic verifiers. Across the four families, we find that credibility falls as availability rises. Regardless of who supplies the judgment, there is no free checker. We then examine what validates a verifier itself, and how much a high score tells us. Three measures appear in the literature: agreement with human labels, the performance of the policy it trains, and behavior under reward hacking. We close with nine metrics that make a verifier claim checkable, and coordinates for the verifiers still to be built.

Several studies showed that Large Language Models (LLMs) can answer medical questions correctly, even outperforming the average human score in some medical exams. However, to our knowledge, no study has been conducted to assess the ability of language models to validate existing or generated medical text for correctness and consistency. In this paper, we introduce MEDEC (https://github.com/abachaa/MEDEC), the first publicly available benchmark for medical error detection and correction in clinical notes, covering five types of errors (Diagnosis, Management, Treatment, Pharmacotherapy, and Causal Organism). MEDEC consists of 3,848 clinical texts, including 488 clinical notes from three US hospital systems that were not previously seen by any LLM. The dataset has been used for the MEDIQA-CORR shared task to evaluate seventeen participating systems [Ben Abacha et al., 2024]. In this paper, we describe the data creation methods and we evaluate recent LLMs (e.g., o1-preview, GPT-4, Claude 3.5 Sonnet, and Gemini 2.0 Flash) for the tasks of detecting and correcting medical errors requiring both medical knowledge and reasoning capabilities. We also conducted a comparative study where two medical doctors performed the same task on the MEDEC test set. The results showed that MEDEC is a sufficiently challenging benchmark to assess the ability of models to validate existing or generated notes and to correct medical errors. We also found that although recent LLMs have a good performance in error detection and correction, they are still outperformed by medical doctors in these tasks. We discuss the potential factors behind this gap, the insights from our experiments, the limitations of current evaluation metrics, and share potential pointers for future research.

Introduction: Recently, the effectiveness of Large Language Models (LLMs) has increased rapidly, allowing them to be used in a great number of applications. However, the risks posed by the generation of false information through LLMs significantly limit their applications in sensitive areas such as healthcare, highlighting the necessity for rigorous validations to determine their utility and reliability. To date, no study has extensively compared the performance of LLMs on Polish medical examinations across a broad spectrum of specialties on a very large dataset. Objectives: This study evaluated the performance of three Generative Pretrained Transformer (GPT) models on the Polish Board Certification Exam (Pa\'nstwowy Egzamin Specjalizacyjny, PES) dataset, which consists of 297 tests. Methods: We developed a software program to download and process PES exams and tested the performance of GPT models using OpenAI Application Programming Interface. Results: Our findings reveal that GPT-3.5 did not pass any of the analyzed exams. In contrast, the GPT-4 models demonstrated the capability to pass the majority of the exams evaluated, with the most recent model, gpt-4-0125, successfully passing 222 (75%) of them. The performance of the GPT models varied significantly, displaying excellence in exams related to certain specialties while completely failing others. Conclusions: The significant progress and impressive performance of LLM models hold great promise for the increased application of AI in the field of medicine in Poland. For instance, this advancement could lead to the development of AI-based medical assistants for healthcare professionals, enhancing the efficiency and accuracy of medical services.

The recent application of LLMs to the legal field has spurred the creation of benchmarks across various jurisdictions and languages. However, no benchmark has yet been specifically designed for the Portuguese legal system. In this work, we present LegalBench.PT, the first comprehensive legal benchmark covering key areas of Portuguese law. To develop LegalBench.PT, we first collect long-form questions and answers from real law exams, and then use GPT-4o to convert them into multiple-choice, true/false, and matching formats. Once generated, the questions are filtered and processed to improve the quality of the dataset. To ensure accuracy and relevance, we validate our approach by having a legal professional review a sample of the generated questions. Although the questions are synthetically generated, we show that their basis in human-created exams and our rigorous filtering and processing methods applied result in a reliable benchmark for assessing LLMs' legal knowledge and reasoning abilities. Finally, we evaluate the performance of leading LLMs on LegalBench.PT and investigate potential biases in GPT-4o's responses. We also assess the performance of Portuguese lawyers on a sample of questions to establish a baseline for model comparison and validate the benchmark.

Large language models can diagnose a student program from its code, problem statement, and reference solution. Evaluating this ability requires a clear definition of what counts as the correct diagnosis. We introduce PyMETA, a Python error dataset with 48,646 student submissions to 155 problems. Every submission has a single label for the first execution error identified by an Online Judge, or No Error when the program passes all tests. A targeted subset of 97 submissions also has expert labels collected through iterative repair and re-execution. The taxonomy has three levels; its most detailed level contains 14 labels, including No Error, Logic Error, named Python exceptions, and an Other Errors category. We evaluate two finetuned models and two groups of prompted LLMs: four earlier models and four recent models. When evaluated against the first execution error, the recent prompted models reach 87.5--93.8% macro F1, above the strongest finetuned baseline at 80.6%. This is the opposite of the comparison obtained with the earlier prompted models. On the 97-item expert subset, however, exact-set match is only 43.3--48.5%, although sample F1 is about 79--81%. Output format also matters. On the same 45 audited submissions whose expert label sets do not contain Logic Error, none of the four recent models returns that label under single-error prompting, but 46.7--57.8% of their multi-error outputs include it, usually after an explicit-error label. The results show that model rankings and claims about label bias depend on the meaning of the gold label, the number of labels a model may return, and the scoring rule.

The AM-DeepSeek-R1-Distilled is a large-scale dataset with thinking traces for general reasoning tasks, composed of high-quality and challenging reasoning problems. These problems are collected from a multitude of open-source datasets, subjected to semantic deduplication and meticulous cleaning to eliminate test set contamination. All responses within the dataset are distilled from reasoning models (predominantly DeepSeek-R1) and have undergone rigorous verification procedures. Mathematical problems are validated by checking against reference answers, code problems are verified using test cases, and other tasks are evaluated with the aid of a reward model. The AM-Distill-Qwen-32B model, which was trained through only simple Supervised Fine-Tuning (SFT) using this batch of data, outperformed the DeepSeek-R1-Distill-Qwen-32B model on four benchmarks: AIME2024, MATH-500, GPQA-Diamond, and LiveCodeBench. Additionally, the AM-Distill-Qwen-72B model surpassed the DeepSeek-R1-Distill-Llama-70B model on all benchmarks as well. We are releasing these 1.4 million problems and their corresponding responses to the research community with the objective of fostering the development of powerful reasoning-oriented Large Language Models (LLMs). The dataset was published in https://huggingface.co/datasets/a-m-team/AM-DeepSeek-R1-Distilled-1.4M{https://huggingface.co/datasets/a-m-team/AM-DeepSeek-R1-Distilled-1.4M}.

Recent large language models (LLMs) have been shown to be effective for misinformation detection. However, the choice of LLMs for experiments varies widely, leading to uncertain conclusions. In particular, GPT-4 is known to be strong in this domain, but it is closed source, potentially expensive, and can show instability between different versions. Meanwhile, alternative LLMs have given mixed results. In this work, we show that Zephyr-7b presents a consistently viable alternative, overcoming key limitations of commonly used approaches like Llama-2 and GPT-3.5. This provides the research community with a solid open-source option and shows open-source models are gradually catching up on this task. We then highlight how GPT-3.5 exhibits unstable performance, such that this very widely used model could provide misleading results in misinformation detection. Finally, we validate new tools including approaches to structured output and the latest version of GPT-4 (Turbo), showing they do not compromise performance, thus unlocking them for future research and potentially enabling more complex pipelines for misinformation mitigation.

Context: Test-driven development (TDD) is a widely employed software development practice that involves developing test cases based on requirements prior to writing the code. Although various methods for automated test case generation have been proposed, they are not specifically tailored for TDD, where requirements instead of code serve as input. Objective: In this paper, we introduce a text-to-testcase generation approach based on a large language model (GPT-3.5) that is fine-tuned on our curated dataset with an effective prompt design. Method: Our approach involves enhancing the capabilities of basic GPT-3.5 for text-to-testcase generation task that is fine-tuned on our curated dataset with an effective prompting design. We evaluated the effectiveness of our approach using a span of five large-scale open-source software projects. Results: Our approach generated 7k test cases for open source projects, achieving 78.5% syntactic correctness, 67.09% requirement alignment, and 61.7% code coverage, which substantially outperforms all other LLMs (basic GPT-3.5, Bloom, and CodeT5). In addition, our ablation study demonstrates the substantial performance improvement of the fine-tuning and prompting components of the GPT-3.5 model. Conclusions: These findings lead us to conclude that fine-tuning and prompting should be considered in the future when building a language model for the text-to-testcase generation task

We present MixtureVitae, an open-access pretraining corpus built to minimize legal risk while providing strong model performance. MixtureVitae follows a risk-mitigated sourcing strategy that combines public-domain and permissively licensed text (e.g., CC-BY/Apache) with carefully justified low-risk additions (e.g., government works and EU TDM-eligible sources), alongside targeted instruction, reasoning and synthetic data with documented provenance. We detail a transparent, multi-stage pipeline for license-aware filtering, safety and quality screening, and domain-aware mixing, and we release the dataset and curation recipes to support reproducible research. In controlled experiments using the open-sci-ref training protocol (fixed architectures at 130M/400M/1.3B/1.7B parameters; training budgets of 50B and 300B tokens), models trained on MixtureVitae consistently outperform other permissive datasets across a suite of standard benchmarks, and at the 1.7B/300B setting they surpass FineWeb-Edu and approach DCLM in the later stages of training. Performance is particularly strong on math/code and competitive on QA tasks. These results demonstrate that permissive-first, risk-mitigated data provides a practical and legally mitigated foundation for training capable LLMs, reducing reliance on indiscriminate web scraping without sacrificing competitiveness. Code: https://github.com/ontocord/mixturevitae

Training language models for formal theorem proving requires large numbers of verifier calls. Slow checks and timeouts often leave GPUs waiting and withhold rewards from potentially correct proofs. We train neural evaluators to predict Lean 4 verdicts and compiler feedback on GPUs using 4.84M labelled programs. We compare two validity classifiers with a feedback model that annotates proof attempts with predicted goal states and error messages. A two-stage evaluator that screens candidates with a classifier and scores its positive predictions with the feedback model reaches 87.8% accuracy in predicting whether Lean accepts a candidate proof, evaluated on a 68,932-program benchmark collected from public provers. The combined evaluator processes 3.5 candidates per second on one H100, 4.2x the throughput of our 64-worker Lean baseline with environment reuse. Finally, we combine our evaluator with Lean verdicts to obtain a hybrid reward system that shortens GRPO training steps by approximately 31%. After one GRPO epoch with hybrid rewards, the prover achieves nearly the same miniF2F pass@1 as with Lean-only rewards (62.2% versus 62.4%) and a pass@32 of 72.4% versus 71.3%.

Deploying Large Language Models (LLMs) for regulatory compliance demands rigorous traceability via comprehensive citations across multi-tiered authority structures. Unlike traditional multi-hop or legal QA, this task requires structured procedural lookups and evidence-set closure rather than entity resolution or case-law reasoning. Existing RAG systems struggle here due to flattened citation edges, fragmented retrieval expansions, and fragile post-hoc attribution. We formalize Regulatory Compliance QA with RegOps-Bench, a novel benchmark featuring an Operational Knowledge Graph derived from complex national R\&D regulations. To address these bottlenecks, we propose RefWalk, a unified framework driven by a shared topic anchor. RefWalk traverses cross-document citations, fuses multi-view candidates via max-based aggregation, and enforces per-rule attribution to explicitly map claims to sources. We establish a strong baseline with substantial improvements in retrieval recall and citation accuracy. Finally, a contrastive evaluation on a U.S. health compliance dataset (HIPAA) reveals that existing systems exhibit saturation on flat-structure rules, underscoring the need for RegOps-Bench. Our code is available at https://github.com/yeongjoonJu/RefWalk.

We describe our development of CSS10, a collection of single speaker speech datasets for ten languages. It is composed of short audio clips from LibriVox audiobooks and their aligned texts. To validate its quality we train two neural text-to-speech models on each dataset. Subsequently, we conduct Mean Opinion Score tests on the synthesized speech samples. We make our datasets, pre-trained models, and test resources publicly available. We hope they will be used for future speech tasks.

We introduce SciVer, the first benchmark specifically designed to evaluate the ability of foundation models to verify claims within a multimodal scientific context. SciVer consists of 3,000 expert-annotated examples over 1,113 scientific papers, covering four subsets, each representing a common reasoning type in multimodal scientific claim verification. To enable fine-grained evaluation, each example includes expert-annotated supporting evidence. We assess the performance of 21 state-of-the-art multimodal foundation models, including o4-mini, Gemini-2.5-Flash, Llama-3.2-Vision, and Qwen2.5-VL. Our experiment reveals a substantial performance gap between these models and human experts on SciVer. Through an in-depth analysis of retrieval-augmented generation (RAG), and human-conducted error evaluations, we identify critical limitations in current open-source models, offering key insights to advance models' comprehension and reasoning in multimodal scientific literature tasks.

We present a benchmark dataset for evaluating method-level code generation task. The benchmark contains a dataset of 175 samples for automated evaluation and a dataset of 161 samples for manual evaluation. We also present a new metric for automatically evaluating the correctness of the generated code, and a set of criteria to manually evaluating the overall quality of the generated code.

Large language models (LLMs) have been widely adopted for clinical question answering (QA). Current systems can attach citations to their answers, but these often point to broad texts, leaving time-pressed clinicians unable to verify them efficiently. An alternative is to ensure that responses are verifiable by construction: providing fine-grained verbatim quotes from reference material that substantiate claims, so users can verify an answer without opening other documents. In this paper, we evaluate the ability of current models to perform this task end-to-end: from providing citations for every factual claim, to producing verbatim quotes, to ensuring that those quotes fully substantiate the claims. To do so, we build a standardized harness over four clinical practice guidelines and evaluate twelve LLMs on 222 synthetic clinical questions, measuring each of these stages separately. We find that most models can attach verbatim quotes to over 90% of their claims from prompting alone, apart from some lightweight models such as claude-haiku-4.5. Yet these quotes often fail to substantiate every detail of the claims they accompany. For instance, claude-opus-5 produces verbatim quotes for 98.0% of its claims, but fully substantiates only 37.1%. Our work provides insights into the current capability gap of LLMs in building verifiable clinical QA systems, along with artifacts for future research.

With the widespread application of Large Language Models (LLMs), their associated security issues have become increasingly prominent, severely constraining their trustworthy deployment in critical domains. This paper proposes a novel safety response framework designed to systematically safeguard LLMs at both the input and output levels. At the input level, the framework employs a supervised fine-tuning-based safety classification model. Through a fine-grained four-tier taxonomy (Safe, Unsafe, Conditionally Safe, Focused Attention), it performs precise risk identification and differentiated handling of user queries, significantly enhancing risk coverage and business scenario adaptability, and achieving a risk recall rate of 99.3%. At the output level, the framework integrates Retrieval-Augmented Generation (RAG) with a specifically fine-tuned interpretation model, ensuring all responses are grounded in a real-time, trustworthy knowledge base. This approach eliminates information fabrication and enables result traceability. Experimental results demonstrate that our proposed safety control model achieves a significantly higher safety score on public safety evaluation benchmarks compared to the baseline model, TinyR1-Safety-8B. Furthermore, on our proprietary high-risk test set, the framework's components attained a perfect 100% safety score, validating their exceptional protective capabilities in complex risk scenarios. This research provides an effective engineering pathway for building high-security, high-trust LLM applications.

As machine learning technology gets applied to actual products and solutions, new challenges have emerged. Models unexpectedly fail to generalize to small changes in the distribution, tend to be confident on novel data they have never seen, or cannot communicate the rationale behind their decisions effectively with the end users. Collectively, we face a trustworthiness issue with the current machine learning technology. This textbook on Trustworthy Machine Learning (TML) covers a theoretical and technical background of four key topics in TML: Out-of-Distribution Generalization, Explainability, Uncertainty Quantification, and Evaluation of Trustworthiness. We discuss important classical and contemporary research papers of the aforementioned fields and uncover and connect their underlying intuitions. The book evolved from the homonymous course at the University of T\"ubingen, first offered in the Winter Semester of 2022/23. It is meant to be a stand-alone product accompanied by code snippets and various pointers to further sources on topics of TML. The dedicated website of the book is https://trustworthyml.io/.

Natural-language software requirements are often ambiguous, inconsistent, and underspecified; in safety-critical domains, these defects propagate into formal models that verify the wrong specification and into implementations that ship unsafe behavior. We show that large language models, equipped with an SMT solver, can audit such requirements: translating them into formal logic, detecting ambiguity through stochastic variation in the generated formalization, and exposing inconsistency, vacuousness, and safety violations through solver queries on the resulting specification. We present VERIMED, a neurosymbolic pipeline that operationalizes this idea for medical-device software requirements, and report two findings. First, stochastic variation across independent formalizations is a signal of ambiguity: requirements that admit multiple plausible interpretations produce SMT-inequivalent formalizations, and bidirectional SMT equivalence checking turns this disagreement into a solver-checkable test. Second, the usefulness of symbolic feedback depends on its granularity: in counterexample-guided repair on a hemodialysis question-answering benchmark, concrete SMT counterexamples raise verified accuracy from 55.4% to 98.5%. Over an extensive experimental evaluation on open-source hemodialysis safety requirements, we show that the LLM-based approach in VERIMED successfully reduces ambiguity-sensitive requirements and enables rigorous auditing of software requirements through SMT-based queries.

In industrial settings, surface defects on steel can significantly compromise its service life and elevate potential safety risks. Traditional defect detection methods predominantly rely on manual inspection, which suffers from low efficiency and high costs. Although automated defect detection approaches based on Convolutional Neural Networks(e.g., Mask R-CNN) have advanced rapidly, their reliability remains challenged due to data annotation uncertainties during deep model training and overfitting issues. These limitations may lead to detection deviations when processing the given new test samples, rendering automated detection processes unreliable. To address this challenge, we first evaluate the detection model's practical performance through calibration data that satisfies the independent and identically distributed (i.i.d) condition with test data. Specifically, we define a loss function for each calibration sample to quantify detection error rates, such as the complement of recall rate and false discovery rate. Subsequently, we derive a statistically rigorous threshold based on a user-defined risk level to identify high-probability defective pixels in test images, thereby constructing prediction sets (e.g., defect regions). This methodology ensures that the expected error rate (mean error rate) on the test set remains strictly bounced by the predefined risk level. Additionally, we observe a negative correlation between the average prediction set size and the risk level on the test set, establishing a statistically rigorous metric for assessing detection model uncertainty. Furthermore, our study demonstrates robust and efficient control over the expected test set error rate across varying calibration-to-test partitioning ratios, validating the method's adaptability and operational effectiveness.

As long-horizon coding agents produce more code than any developer can review, oversight collapses onto a single surface: the automated test suite. Reward hacking naturally arises in this setup, as the agent optimizes for passing tests while deviating from the users true goal. We study this reward hacking phenomenon by decompose software engineering tasks into three parts: (i) a natural language description of the specification (ii) visible validation tests that exercise specified features in isolation, and (iii) held-out tests that compose those same features to simulate real-world usage. Based on the specification and the visible validation test suites, a genuine agent would be able to generate a solution that can also pass all of the held-out tests. Therefore we use the gap in pass rates on these two suites to quantify reward hacking. Based on this methodology, we introduce SpecBench, a benchmark comprising 30 systems-level programming tasks ranging from short horizon tasks like building a JSON parser to ultra long horizon tasks like building an entire OS kernel from scratch. Large-scale experiments reveal a consistent pattern: while every frontier agent saturates the visible suite, reward hacking persists, with smaller models exhibiting larger gaps on holdout suites. The gap also scales sharply with task length: it grows by 28 percentage points for every tenfold increase in code size. Failures range from subtle feature isolation to deliberate exploits, including a 2,900-line hash-table "compiler" that memorizes test inputs. SpecBench offers a principled testbed for measuring whether coding agents build genuine working systems or merely game the test suites developers hand them.

The rapid development in large language models (LLMs) has transformed the landscape of natural language processing and understanding (NLP/NLU), offering significant benefits across various domains. However, when applied to scientific research, these powerful models exhibit critical failure modes related to scientific integrity and trustworthiness. Existing general-purpose LLM guardrails are insufficient to address these unique challenges in the scientific domain. We provide comprehensive guidelines for deploying LLM guardrails in the scientific domain. We identify specific challenges -- including time sensitivity, knowledge contextualization, conflict resolution, and intellectual property concerns -- and propose a guideline framework for the guardrails that can align with scientific needs. These guardrail dimensions include trustworthiness, ethics & bias, safety, and legal aspects. We also outline in detail the implementation strategies that employ white-box, black-box, and gray-box methodologies that can be enforced within scientific contexts.

Financial document recognition is often assessed using character and word error rates, although these metrics do not express whether a system preserves the financial facts on which people and organisations act. This paper develops a human-centred, risk-constrained framework for evaluating financial document recognition as evidence recovery and decision support. It defines typed financial facts linked to visible source regions and separates verbatim transcription, deterministic normalization, and business interpretation. A layered quality model covers text, localization, layout, tables, fields, fact correctness, evidence grounding, uncertainty, robustness, and operational performance. Non-compensatory acceptance gates use one-sided uncertainty bounds for critical-field errors, preventing strong aggregate scores from masking failures in amounts, identifiers, signs, periods, or entity associations. Selective automation is formulated as a constrained routing problem that accounts for residual risk, human-review capacity, and review cost. The framework further specifies annotation and adjudication procedures, leakage-resistant benchmark partitions, statistical estimands, subgroup analyses, and a reproducibility card. Formal results establish the limitations of compensatory scoring, the non-zero upper risk bound after zero observed errors, and a loss sensitive decision rule for routing uncertain cases to human review. The contribution is theoretical and methodological; no new model, dataset, or empirical performance result is claimed.

With the rapid evolution of large language models (LLMs), new and hard-to-predict harmful capabilities are emerging. This requires developers to be able to identify risks through the evaluation of "dangerous capabilities" in order to responsibly deploy LLMs. In this work, we collect the first open-source dataset to evaluate safeguards in LLMs, and deploy safer open-source LLMs at a low cost. Our dataset is curated and filtered to consist only of instructions that responsible language models should not follow. We annotate and assess the responses of six popular LLMs to these instructions. Based on our annotation, we proceed to train several BERT-like classifiers, and find that these small classifiers can achieve results that are comparable with GPT-4 on automatic safety evaluation. Warning: this paper contains example data that may be offensive, harmful, or biased.

Statutes are increasingly parsed by machines before people read them, and the parsers disagree: on Missouri's statutes, two independently written extractors diverge on numeric-threshold presence at a false-negative rate of 0.43. We ask what formal logic survives such noise. We build a passive survival certificate for the Duquenne-Guigues implication basis of machine-extracted statutory contexts: per-attribute inter-extractor disagreement is measured, replayed against the basis in 1,000 Monte Carlo trials, and an implication is certified only when a one-sided Wilson 95% lower bound on survival reaches 0.95; every certified implication carries premise spans and a minimal counterexample. On 29,365 Missouri sections and 502 Indian central-Act sections, the preregistered held-out gate passes (10 statute families across 7 Titles exact; 16 across 11 with 5% tolerance), yet under one globally deployed error model 93.2% of held-out chapters fall below the informativeness floor, and a 2x2 factorial assigns that to calibration-rate transfer, not selection. The certificate is usable but fragile: deploy it per-chapter-calibrated or error-tolerant. Code, data products, and the audit trail, including one retracted claim, are released.

Due to the cumbersome nature of human evaluation and limitations of code-based evaluation, Large Language Models (LLMs) are increasingly being used to assist humans in evaluating LLM outputs. Yet LLM-generated evaluators simply inherit all the problems of the LLMs they evaluate, requiring further human validation. We present a mixed-initiative approach to ``validate the validators'' -- aligning LLM-generated evaluation functions (be it prompts or code) with human requirements. Our interface, EvalGen, provides automated assistance to users in generating evaluation criteria and implementing assertions. While generating candidate implementations (Python functions, LLM grader prompts), EvalGen asks humans to grade a subset of LLM outputs; this feedback is used to select implementations that better align with user grades. A qualitative study finds overall support for EvalGen but underscores the subjectivity and iterative process of alignment. In particular, we identify a phenomenon we dub criteria drift: users need criteria to grade outputs, but grading outputs helps users define criteria. What is more, some criteria appears dependent on the specific LLM outputs observed (rather than independent criteria that can be defined a priori), raising serious questions for approaches that assume the independence of evaluation from observation of model outputs. We present our interface and implementation details, a comparison of our algorithm with a baseline approach, and implications for the design of future LLM evaluation assistants.

Scientific images function as critical evidence in research communication, yet their integrity faces unprecedented threats from AI-generated content that introduces subtle but consequential errors. Existing evaluation paradigms prove inadequate: perceptual quality metrics poorly correlate with scientific validity, while language models lack domain-specific verification capabilities. To address this gap, we propose the Scientific Image Utility and Upgradability Assessment (SIU^2A) framework, which introduces two complementary dimensions for scientific image evaluation. Utility encompasses error detection (identifying scientific inaccuracies) and correction feasibility (assessing whether errors can be reliably repaired). Upgradability measures the quality of correction. We categorize scientific image corruption into four fundamental types: Detail Distortion, Incompleteness, False Content, and Entity Confusion. Based on this taxonomy, we construct SIU^2A-Benchmark, a dataset with expert annotations for error identification and repair. The framework implements a two-stage evaluation protocol: the Utility stage evaluates error detection capability and repair instruction generation, while the Upgradability stage assesses whether corrections faithfully restore scientific validity without compromising existing accurate information. Experiments reveal that current multimodal systems exhibit significant limitations in both scientific error assessment and faithful correction, exposing a fundamental gap between visual perception and scientific usability.

Monitoring complex industrial assets relies on engineer-authored symbolic rules that trigger based on sensor conditions and prompt technicians to perform corrective actions. The bottleneck is not detection but response: translating rules into maintenance steps requires asset-specific knowledge gained through years of practice. We investigate whether LLMs can serve as decision support for this rule-to-action step and introduce , a benchmark of 6{,}690 expert-validated multiple-choice questions from 118 rule-action pairs across 16 asset types. We contribute (i) a symbolic-to-MCQA pipeline normalizing rules to Disjunctive Normal Form with embedding-based distractor sampling, (ii) five variants probing distinct failure modes (Pro, Pert, Verbose, Aug, Rationale), and (iii) a benchmark of 29 LLMs and 4 embedding baselines. A human evaluation (9 practitioners, mean 45.0\%) confirms requires specialist knowledge beyond operational experience. Three findings stand out. The frontier has closed: the top three LLMs lie within one Macro point, with Bradley-Terry Elo placing claude-opus-4-6 30 points above the next model. Yet \,Pro exposes brittleness, with every model losing 13--60\% relative accuracy under distractor expansion. \,Aug exposes pattern-matching: under condition inversion, frontier models still select the original answer 49--63\% of the time. The deployment bottleneck is not capability but calibration: frontier models handle template-style fault detection but break under structural perturbation.

Ultrasound is crucial in modern medicine but faces challenges like operator dependence, image noise, and real-time scanning, hindering AI integration. While large multimodal models excel in other medical imaging areas, they struggle with ultrasound's complexities. To address this, we introduce Dolphin v1.0 (V1) and its reasoning-augmented version, Dolphin R1-the first large-scale multimodal ultrasound foundation models unifying diverse clinical tasks in a single vision-language framework.To tackle ultrasound variability and noise, we curated a 2-million-scale multimodal dataset, combining textbook knowledge, public data, synthetic samples, and general corpora. This ensures robust perception, generalization, and clinical adaptability.The Dolphin series employs a three-stage training strategy: domain-specialized pretraining, instruction-driven alignment, and reinforcement-based refinement. Dolphin v1.0 delivers reliable performance in classification, detection, regression, and report generation. Dolphin R1 enhances diagnostic inference, reasoning transparency, and interpretability through reinforcement learning with ultrasound-specific rewards.Evaluated on U2-Bench across eight ultrasound tasks, Dolphin R1 achieves a U2-score of 0.5835-over twice the second-best model (0.2968) setting a new state of the art. Dolphin v1.0 also performs competitively, validating the unified framework. Comparisons show reasoning-enhanced training significantly improves diagnostic accuracy, consistency, and interpretability, highlighting its importance for high-stakes medical AI.

As large language models (LLMs) become more powerful and are deployed more autonomously, it will be increasingly important to prevent them from causing harmful outcomes. Researchers have investigated a variety of safety techniques for this purpose, e.g. using models to review the outputs of other models, or red-teaming techniques to surface subtle failure modes. However, researchers have not evaluated whether such techniques still ensure safety if the model is itself intentionally trying to subvert them. In this paper, we develop and evaluate pipelines of safety techniques ("protocols") that are robust to intentional subversion. We investigate a scenario in which we want to solve a sequence of programming problems, using access to a powerful but untrusted model (in our case, GPT-4), access to a less powerful trusted model (in our case, GPT-3.5), and limited access to high-quality trusted labor. We investigate protocols that aim to never submit solutions containing backdoors, which we operationalize here as logical errors that are not caught by test cases. We investigate a range of protocols and test each against strategies that the untrusted model could use to subvert them. One protocol is what we call trusted editing. This protocol first asks GPT-4 to write code, and then asks GPT-3.5 to rate the suspiciousness of that code. If the code is below some suspiciousness threshold, it is submitted. Otherwise, GPT-3.5 edits the solution to remove parts that seem suspicious and then submits the edited code. Another protocol is untrusted monitoring. This protocol asks GPT-4 to write code, and then asks another instance of GPT-4 whether the code is backdoored, using various techniques to prevent the GPT-4 instances from colluding. These protocols improve substantially on simple baselines.

Given the increasing adoption of modern AI-enabled control systems, ensuring their safety and reliability has become a critical task in software testing. One prevalent approach to testing control systems is falsification, which aims to find an input signal that causes the control system to violate a formal safety specification using optimization algorithms. However, applying falsification to AI-enabled control systems poses two significant challenges: (1)~it requires the system to execute numerous candidate test inputs, which can be time-consuming, particularly for systems with AI models that have many parameters, and (2)~multiple safety requirements are typically defined as a conjunctive specification, which is difficult for existing falsification approaches to comprehensively cover. This paper introduces Synthify, a falsification framework tailored for AI-enabled control systems. Our approach performs falsification in a two-phase process. At the start, Synthify synthesizes a program that implements one or a few linear controllers to serve as a proxy for the AI controller. This proxy program mimics the AI controller's functionality but is computationally more efficient. Then, Synthify employs the ε-greedy strategy to sample a promising sub-specification from the conjunctive safety specification. It then uses a Simulated Annealing-based falsification algorithm to find violations of the sampled sub-specification for the control system. To evaluate Synthify, we compare it to PSY-TaLiRo, a state-of-the-art and industrial-strength falsification tool, on 8 publicly available control systems. On average, Synthify achieves a 83.5% higher success rate in falsification compared to PSY-TaLiRo with the same budget of falsification trials. The safety violations found by Synthify are also more diverse than those found by PSY-TaLiRo, covering 137.7% more sub-specifications.

Large language models can generate plausible code-review comments, but such comments may contain technically incorrect claims that mislead developers. We study technical trustworthiness judgment: determining whether a review comment's core technical claims are correct and applicable to the reviewed code in its repository context. Existing code-review benchmarks primarily evaluate review generation, issue discovery, or general comment quality, but do not directly assess whether an agent can determine the technical trustworthy of an individual review comment. To fill this gap, we introduce CRJudgeBench, a benchmark of 1199 instances constructed from real pull requests and expert-verified perturbations, covering both trustworthy and plausible but untrustworthy comments. We further present Sentinel, a repository-grounded agentic judge that actively gathers code evidence to verify review comments before making judgments. Starting from Qwen3-Coder-30B-A3B-Instruct, Sentinel is trained on the CRJudgeBench training split through iterative action-level learning from a privileged teacher. On the 359-instance CRJudgeBench test set, Sentinel achieves 76.60\% accuracy, outperforming GLM-5.3 by 6.13 percentage points and its base model by 19.78 points. These results show that even state-of-the-art general-purpose LLMs struggle to identify untrustworthy comments, while iterative action-level learning substantially improves the accuracy of repository-grounded trustworthiness judgments. Our dataset is available at https://huggingface.co/datasets/dcloud347/CRJudgeBenchmark

The rapid advancement of artificial intelligence (AI) systems in critical domains like healthcare, justice, and social services has sparked numerous regulatory initiatives aimed at ensuring their safe deployment. Current regulatory frameworks, exemplified by recent US and EU efforts, primarily focus on procedural guidelines while presuming that scientific benchmarking can effectively validate AI safety, similar to how crash tests verify vehicle safety or clinical trials validate drug efficacy. However, this approach fundamentally misunderstands the unique technical challenges posed by modern AI systems. Through systematic analysis of successful technology regulation case studies, we demonstrate that effective scientific regulation requires a causal theory linking observable test outcomes to future performance - for instance, how a vehicle's crash resistance at one speed predicts its safety at lower speeds. We show that deep learning models, which learn complex statistical patterns from training data without explicit causal mechanisms, preclude such guarantees. This limitation renders traditional regulatory approaches inadequate for ensuring AI safety. Moving forward, we call for regulators to reckon with this limitation, and propose a preliminary two-tiered regulatory framework that acknowledges these constraints: mandating human oversight for high-risk applications while developing appropriate risk communication strategies for lower-risk uses. Our findings highlight the urgent need to reconsider fundamental assumptions in AI regulation and suggest a concrete path forward for policymakers and researchers.

Chain-of-thought reasoning has demonstrated significant improvements in large language models and vision language models, yet its potential for audio language models remains largely unexplored. In this technical report, we take a preliminary step towards closing this gap. For better assessment of sound reasoning, we propose AF-Reasoning-Eval, a benchmark targeting common-sense reasoning and the ability to discriminate among closely related choices. To prepare training corpus for sound reasoning abilities, we propose automatic pipelines that transform existing audio question answering and classification data into explicit reasoning chains, yielding AF-CoT-Train with 1.24M samples. We study the effect of finetuning Audio Flamingo series on AF-CoT-Train and observe considerable improvements on several reasoning benchmarks, validating the effectiveness of chain-of-thought finetuning on advanced sound understanding.

We present SpaceQA, to the best of our knowledge the first open-domain QA system in Space mission design. SpaceQA is part of an initiative by the European Space Agency (ESA) to facilitate the access, sharing and reuse of information about Space mission design within the agency and with the public. We adopt a state-of-the-art architecture consisting of a dense retriever and a neural reader and opt for an approach based on transfer learning rather than fine-tuning due to the lack of domain-specific annotated data. Our evaluation on a test set produced by ESA is largely consistent with the results originally reported by the evaluated retrievers and confirms the need of fine tuning for reading comprehension. As of writing this paper, ESA is piloting SpaceQA internally.

Enterprise large language model (LLM) applications often begin as prototypes whose behavior is carried by prompts and retrieval context. Productization adds requirements for source boundaries, entity routing, answer contracts, and reproducible traces. We present a harness-engineering approach that reconstructs this pattern into a traceable, auditable LLM-agent architecture: deterministic behavior moves into code, manifests, schemas, and validation artifacts around a replaceable composition boundary, while source-backed claims remain the authority for runtime answers. We instantiate it on a public-data slice of five Korean corporate groups (25 listed companies) and evaluate three research questions. (1) The harness preserves its source-grounding, entity-routing, trace, output-hygiene, and recommendation-language contracts across the fixed validation scenarios; a fault-injection control confirms the validators flag deliberately broken contracts. (2) The checks the harness enforces held under model substitution: across three hosted models, they passed on all 270 composition-boundary runs; failures were confined to the model-composed side and were caught and recorded. (3) The code-owned guarantees are load-bearing, not reproducible by prompting alone: holding the model fixed and varying only the enforcement layer, prompt instructions alone let recommendation-language and internal-trace-leakage violations reach the reader, which the harness blocks entirely. A bolt-on external guardrail prevents such violations too but over-refuses, dropping utility to 88/120 where the harness preserves full utility (120/120); in this ablation, only code-owned enforcement preserves both safety and utility. The result is a reusable engineering pattern for turning exploratory prototypes into auditable applications with versioned source, control, and validation artifacts.