LightOnOCR-3 🦉 Collection High-Performance OCR and Layout Extraction in One Model • 4 items • Updated 18 days ago • 22
OpenJev Collection 27B decision model. Text or a screenshot plus your options in, a scored decision out, in one pass. • 6 items • Updated 19 days ago • 1
OpenJev Flash 9B Collection The 9B decision model. About 1.6x faster than OpenJev, for text decisions. • 6 items • Updated 19 days ago • 1
MergeMix: Optimizing Mid-Training Data Mixtures via Learnable Model Merging Paper • 2601.17858 • Published Jan 25 • 2
Bounding Hallucinations: Information-Theoretic Guarantees for RAG Systems via Merlin-Arthur Protocols Paper • 2512.11614 • Published Jan 30 • 1
view article Article The Open Source Community is backing OpenEnv for Agentic RL +18 burtenshaw, spisakjo, lysandre, darktex, willcb, qjoy, pawalt, cwing-nv, danielhanchen, andrewzhou, thegovind, shimmyshimmer, Hamid-Nazeri, Sanyam, zkwentz, emre0, lewtun, sergiopaniego, banghua, unseenmars • Jun 8 • 123
The Bestiary Collection Decensored language models made using Heretic (https://github.com/p-e-w/heretic) • 5 items • Updated May 21 • 123
Borealis — training datasets Collection Audio + text datasets used to train Vikhrmodels/Borealis-5b-it (Russian audio-LLM). • 8 items • Updated May 12 • 2
Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference Paper • 2412.13663 • Published Dec 18, 2024 • 173
Granite Libraries Collection Collections of adapters that extend Granite models. • 4 items • Updated 14 days ago • 18
LightMem: Lightweight and Efficient Memory-Augmented Generation Paper • 2510.18866 • Published Oct 21, 2025 • 116
SuperLocalMemory V3: Information-Geometric Foundations for Zero-LLM Enterprise Agent Memory Paper • 2603.14588 • Published Mar 15 • 4
The Missing Knowledge Layer in Cognitive Architectures for AI Agents Paper • 2604.11364 • Published Apr 13 • 2
Human-Like Lifelong Memory: A Neuroscience-Grounded Architecture for Infinite Interaction Paper • 2603.29023 • Published Mar 30 • 2
Granite Guardian Collection Safety models for detecting risks, toxicity, and hallucinations in LLM workflows. • 15 items • Updated 14 days ago • 30