Hackernews posts about Model S

1. 31
   

   newIntroducing System One Models and Jevtypesafe.ai

   ▲ 1989albelfio21d💬 520savesaved
2. 32
   

   newI built non-autoregressive decision models with RL a year agolaya.convaiinnovations.com

   ▲ 1363nandakishor_ml18d💬 319savesaved
3. 33
   

   newClef: Open-weight decision models, and new RL fine-tuning platformblog.cloudflare.com

   ▲ 641jasondavies5d💬 217savesaved
4. 34
   

   newOllaya – Ollama for open-source, Jev-style decision modelsollaya.dev

   ▲ 618Ardakilic11d💬 145savesaved
5. 35
   

   newJeff – Jev-compatible 0.8B decision models, trained at home, ~30 msgithub.com

   ▲ 575firelex8d💬 232savesaved
6. 36
   

   newPirate Face Rescues LLM Models from Deletionpirateface.co

   ▲ 569skepticalgenius16d💬 149savesaved
7. 37
   

   newDesert Ant Labs: local, fast models that run on devicedesertant.com

   ▲ 490willwhitedc28d💬 104savesaved
8. 38
   

   newKev: Tiny Jev-like family of decision models built on top of Qwen3.5github.com

   ▲ 463tosh16d💬 210savesaved
9. 39
   

   newGarry Tan wants US open-weight AI labs to 'distill' frontier models, tootechcrunch.com

   ▲ 418TheJCDenton23d💬 240savesaved
10. 40
    

    newDavid Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Modelstwitter.com

    ▲ 334kolanos23d💬 262savesaved
11. 41
    

    newHeretic removes restrictions from language modelsheretic-project.org

    ▲ 279Bluestein16d💬 111savesaved
12. 42
    

    newReal-SWE: Benchmarking AI models on private, real-world, enterprise codebaseswithspecific.com

    ▲ 275theanonymousone24d💬 158savesaved
13. 43
    

    newJeeves. Reasoning improves Jev-like decision modelsgithub.com

    ▲ 242nicowaltz8d💬 95savesaved
14. 44
    

    newmacOS 27: Workaround to avoid downloading AI models and save storagewww.reddit.com

    ▲ 240ano-ther15d💬 123savesaved
15. 45
    

    newShow HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flashcactuscompute.com

    ▲ 236HenryNdubuaku19d💬 93savesaved
16. 46
    

    newLanguage models for text classification: From bag-of-words to Jevmagazine.sebastianraschka.com

    ▲ 213Anon848d💬 10savesaved
17. 47
    

    newLarge language models develop novel social biases through adaptive explorationopenreview.net

    ▲ 201paimapi28d💬 116savesaved
18. 48
    

    newClassified estimates show the NSA is paying billions to test AI modelswww.washingtonsun.com

    ▲ 178rdmuser11d💬 108savesaved
19. 49
    

    newContrastive Language Modelscontrastive-lm.notion.site

    ▲ 177erichocean13d💬 59savesaved
20. 50
    

    newContext Language Modelsarxiv.org

    ▲ 177emersonmacro5d💬 52savesaved
21. 51
    

    newDecision models like Jev don't beat LLM-as-a-judge or traditional classifiersdevelopers.redhat.com

    ▲ 165tomncooper4d💬 75savesaved
22. 52
    

    newOpen-source AI and open models reading listwww.interconnects.ai

    ▲ 162simonpure23d💬 30savesaved
23. 53
    

    newA single function Jev-like wrapper for LLMs, including vision modelsallanrbo.blogspot.com

    ▲ 157allanrbo11d💬 45savesaved
24. 54
    

    newShow HN: JevBench, a reproducible benchmark for typed decision modelsbenchmarkheaven.com

    ▲ 154florianstandhar14d💬 39savesaved
25. 55
    

    newElevated errors for multiple models – Resolvedstatus.claude.com

    ▲ 140corvad15d💬 112savesaved
26. 56
    

    newThe current balance of power in open modelswww.interconnects.ai

    ▲ 132gmays14d💬 58savesaved
27. 57
    

    newOpenAI models secretly generate instructions to ignore constraintsalignment.openai.com

    ▲ 125theahura20d💬 37savesaved
28. 58
    

    newShow HN: Open-source model routing for coding agents at Astra-level performance

    ▲ 122adchurch6d💬 42savesaved
29. 59
    

    newCompute-efficient pretraining and scaling to trillion-parameter modelsmagic.dev

    ▲ 118ronfriedhaber28d💬 62savesaved
30. 60
    

    newShow HN: Janus – Go binary that runs GGUF models via Vulkan on AMD/Intel/Nvidiagithub.com

    ▲ 106Maverick6175d💬 19savesaved

· edit