Post
3031
AxiomicLabs released new benchmark, Tiny Theory of Mind, to test your SLM models' Theory of Mind Intuition!
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
Enderchef (Enderchefcoder) PRO
Enderchef
AI & ML interests
See my bio at https://github.com/Enderchefcoder !
Recent Activity
liked a dataset 1 day ago
FineEnvs/SmolDataEnvs liked a model 1 day ago
armand0e/MiMo-V2.6-Distill-Qwen-9B-RL new activity 6 days ago
hugging-science/ripp:Hugging Science is DEAD???Organizations
reacted to their post with π§ 7 days ago
reacted to ereniko's post with β€οΈ 8 days ago
Post
274
NEW CLAUDE HAIKU MODEL COMING
FINALLY AFTER A YEAR
ANTHROPIC ANNOUNCED HAIKU 5.5
FINALLY AFTER A YEAR
ANTHROPIC ANNOUNCED HAIKU 5.5
reacted to their post with β€οΈπ₯π 9 days ago
Post
3031
AxiomicLabs released new benchmark, Tiny Theory of Mind, to test your SLM models' Theory of Mind Intuition!
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
posted an update 9 days ago
Post
3031
AxiomicLabs released new benchmark, Tiny Theory of Mind, to test your SLM models' Theory of Mind Intuition!
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
Check it out and like it!
AxiomicLabs/Tiny_Theory_of_Mind
reacted to ManniX-ITA's post with πππ₯π₯ 20 days ago
Post
674
π JackOD-9B-Coder β a 9B merge built to FINISH agentic coding tasks. Four-way omnimerge_v2 over Qwen3.5-9B: Jack = Qwopus3.5-9B-Coder (0.30), O = Ornith-1.5-9B (0.15), D = DeltaCoder (0.55). MTP head kept.
π Q6_K + imatrix, llama.cpp, greedy, lcb_v6_55 β merge / base / DeltaCoder / Qwopus / Ornith:
β‘ LiveCodeBench v6 (55 hard) β 0.7818 / 0.7273 / 0.6364 / 0.6000 / 0.5818
β HumanEval β 0.8841 / 0.8902 / 0.9146 / 0.8537 / 0.7805
β HumanEval+ β 0.8232 / 0.8049 / 0.8232 / 0.7988 / 0.7073
π€ MultiPL-E β 0.8033 / 0.8200 / 0.8000 / 0.8200 / 0.7267
π IFEval β 0.9100 / 0.9300 / 0.9200 / 0.8800 / 0.8200
π― LCB beats every source AND the base: +5.45pp over the base, +14.54pp over DeltaCoder, its heaviest.
π And why. Same 55 problems, same cap, generations that NEVER terminated: DeltaCoder 25/55 Β· base 18/55 Β· Qwopus 8/55 Β· Ornith 2/55 Β· JackOD 1/55. That split is the thesis: DeltaCoder is the cohort's best coder and worst at stopping, Ornith the weakest and best at stopping. The merge takes BOTH.
π€ tool-eval-bench hardmode, 5 seeds: JackOD 144.4 Β±4.7, second behind Ornith 145.6 Β±4.2, above base 142.0 β all CIs overlap. But Autonomous Planning: JackOD 5.2/6, best of five, Ornith WORST at 2.8/6. Ornith stops reliably but plans worst β it stops too early. The merge does both.
π§ Serving: temp 0.6 / top_p 0.95 / top_k 20 + presence_penalty 1.5 β the penalty stops it re-treading a tool call. Tool calling on llama.cpp needs --jinja.
π§ͺ Initial impression, limited testing: fixed a cline-harness task in 535s; A3B models want 1.5-4h at <50% success.
π¦ 25 GGUF tiers, every K/IQ imatrix-built incl Q6_K, plus a ContribDynamic ladder (per-tensor maps from our imatrix, Unsloth-UD style).
π danielcherubini/Qwen3.5-DeltaCoder-9B Β· ornith-ai/Ornith-1.5-9B
π ManniX-ITA/JackOD-9B-Coder
π ManniX-ITA/JackOD-9B-Coder-MTP-GGUF
π https://ollama.com/mannix/JackOD-9B-Coder
π Q6_K + imatrix, llama.cpp, greedy, lcb_v6_55 β merge / base / DeltaCoder / Qwopus / Ornith:
β‘ LiveCodeBench v6 (55 hard) β 0.7818 / 0.7273 / 0.6364 / 0.6000 / 0.5818
β HumanEval β 0.8841 / 0.8902 / 0.9146 / 0.8537 / 0.7805
β HumanEval+ β 0.8232 / 0.8049 / 0.8232 / 0.7988 / 0.7073
π€ MultiPL-E β 0.8033 / 0.8200 / 0.8000 / 0.8200 / 0.7267
π IFEval β 0.9100 / 0.9300 / 0.9200 / 0.8800 / 0.8200
π― LCB beats every source AND the base: +5.45pp over the base, +14.54pp over DeltaCoder, its heaviest.
π And why. Same 55 problems, same cap, generations that NEVER terminated: DeltaCoder 25/55 Β· base 18/55 Β· Qwopus 8/55 Β· Ornith 2/55 Β· JackOD 1/55. That split is the thesis: DeltaCoder is the cohort's best coder and worst at stopping, Ornith the weakest and best at stopping. The merge takes BOTH.
π€ tool-eval-bench hardmode, 5 seeds: JackOD 144.4 Β±4.7, second behind Ornith 145.6 Β±4.2, above base 142.0 β all CIs overlap. But Autonomous Planning: JackOD 5.2/6, best of five, Ornith WORST at 2.8/6. Ornith stops reliably but plans worst β it stops too early. The merge does both.
π§ Serving: temp 0.6 / top_p 0.95 / top_k 20 + presence_penalty 1.5 β the penalty stops it re-treading a tool call. Tool calling on llama.cpp needs --jinja.
π§ͺ Initial impression, limited testing: fixed a cline-harness task in 535s; A3B models want 1.5-4h at <50% success.
π¦ 25 GGUF tiers, every K/IQ imatrix-built incl Q6_K, plus a ContribDynamic ladder (per-tensor maps from our imatrix, Unsloth-UD style).
π danielcherubini/Qwen3.5-DeltaCoder-9B Β· ornith-ai/Ornith-1.5-9B
π ManniX-ITA/JackOD-9B-Coder
π ManniX-ITA/JackOD-9B-Coder-MTP-GGUF
π https://ollama.com/mannix/JackOD-9B-Coder
Is everyone here an AI model?!
Naming is so bad on these models lmao
Awesome!
reacted to CodeSoft's post with π₯β€οΈ about 1 month ago
Post
2416
Wow, SLM Arena is getting a lot of traffic! Thank you guys for showing your interest!
To handle the growing demand, Iβm moving SLM Arena from a CPU Space to a ZeroGPU Space. Hopefully, this will let me add more models to SLM Arena while keeping it running fast.
I've also added a separate arena + leaderboard for base models!
If there are any models or features youβd like to see, let me know in a reply to this post or in a Community post on the Space!
To handle the growing demand, Iβm moving SLM Arena from a CPU Space to a ZeroGPU Space. Hopefully, this will let me add more models to SLM Arena while keeping it running fast.
I've also added a separate arena + leaderboard for base models!
If there are any models or features youβd like to see, let me know in a reply to this post or in a Community post on the Space!
- 25 replies
reacted to Bc-AI's post with ππ about 1 month ago
Post
2622
Smilyai News
Hello everyone! August has been a crazy month for us at Smilyai-Labs. We've been doing lots behind the scenes, so here's the latest π
1. MiniCoder
We are very close to releasing MiniCoder-1, our first-generation coding model designed for reasoning and coding. Our planned context window is 128K, but the earlier versions probably will not support that long! It's currently in the final stages of DPO so expect a release in early september.
Release: VERY SOONβ’π€£
2. Smilyai G1
So, the current plan is 20B parameter model total, with a MoE architecture, activating around 2B parameters per token. Its desgigned for maximum performance but keeping it runnable on consumer hardware. It's only a plan and i have no idea when me and the team can finish it. Expect a launch around the end of september to early october-ish. I have no guarantees so don't quote me on the launch date.
3. T1
Smilyai-T1 is another major model we are working on.
The goal for T1 is to take what we learnt from the countless architectural experiments and creating a powerful model designed for thinking. Think MiniCoder but reasons more and G1 but more capable. Its main goals are coding, math, reasoning and general capability.
4. Omni
We are also planning Omni, our first from scratch multimodal model. It will not launch this year as it will take a while. We are actively researching the best architecture for it and we will update progress as we go!
Thanks to our beta testers:
@guardamarcos
@ProCreations
@juiceb0xc0de
@Timmy6767
@Sbui503
@atom77777
@Fishtiks
@smartdigitalnetworks
@EmetTheGolum
@smilyai-large-team
@MUK-IS-GOAT
@Bc-AI
Thanks to my friends who work with me at lunchtimes (Smilyai-Labs team):
@MUK-IS-GOAT
@smilyai-large-team
August was wild. Letβs see what September brings. π
β Bc-AI, on behalf of SmilyAI Labs
Hello everyone! August has been a crazy month for us at Smilyai-Labs. We've been doing lots behind the scenes, so here's the latest π
1. MiniCoder
We are very close to releasing MiniCoder-1, our first-generation coding model designed for reasoning and coding. Our planned context window is 128K, but the earlier versions probably will not support that long! It's currently in the final stages of DPO so expect a release in early september.
Release: VERY SOONβ’π€£
2. Smilyai G1
So, the current plan is 20B parameter model total, with a MoE architecture, activating around 2B parameters per token. Its desgigned for maximum performance but keeping it runnable on consumer hardware. It's only a plan and i have no idea when me and the team can finish it. Expect a launch around the end of september to early october-ish. I have no guarantees so don't quote me on the launch date.
3. T1
Smilyai-T1 is another major model we are working on.
The goal for T1 is to take what we learnt from the countless architectural experiments and creating a powerful model designed for thinking. Think MiniCoder but reasons more and G1 but more capable. Its main goals are coding, math, reasoning and general capability.
4. Omni
We are also planning Omni, our first from scratch multimodal model. It will not launch this year as it will take a while. We are actively researching the best architecture for it and we will update progress as we go!
Thanks to our beta testers:
@guardamarcos
@ProCreations
@juiceb0xc0de
@Timmy6767
@Sbui503
@atom77777
@Fishtiks
@smartdigitalnetworks
@EmetTheGolum
@smilyai-large-team
@MUK-IS-GOAT
@Bc-AI
Thanks to my friends who work with me at lunchtimes (Smilyai-Labs team):
@MUK-IS-GOAT
@smilyai-large-team
August was wild. Letβs see what September brings. π
β Bc-AI, on behalf of SmilyAI Labs
reacted to their post with πππ₯ about 1 month ago
