webAI ships 3.66B formal-logic model TwIL-LM3-Pro, 2.09 GiB and runs locally
kimmonismus · x · 2026-10-02
webAI released TwIL-LM3-Pro, a 3.66B model specialized for formal logic, built by post-training IBM's Granite 4.2 to check whether conclusions follow from premises, translate statements into logic, and critique formal proofs.
- Roughly matches Qwen3-8B on the headline formal-logic composite with less than half the parameters
- Beats the public VibeThinker-3B checkpoint on all six reported formal tasks
- Q4 GGUF weights are just 2.09 GiB and run locally via llama.cpp
- Weights and model card available under a non-commercial license
More from Models
- Local benchmark: community 9B model shows 4x lower latency on RTX 5070 vs peers — Storge2 · 2026-10-02
- Fable 5.5 release 'impending, as soon as next week', claims Bindu Reddy — bindureddy · 2026-10-02
- Open weights will be good enough: why frontier AI smarts won't matter for daily life — Dan_Jeffries1 · 2026-10-02
- Open-weight models trail frontier by just 4 months — here's when to use them — TechPreacher · 2026-10-02
- MiniMax M3.1 Flash joins RSI Arena mid-run as compute competition heats up — my_cat_can_code · 2026-10-02
- Morgan Stanley's Parallel Power Tempering sampling rivals RL post-training without weight updates — morganstanley · 2026-10-02