Tilelli LLM:采用每 Token 路由机制的开源反幻觉模型

themoroccanship · reddit · 2026-08-06

Tilelli Lab has released Tilelli LLM, their first open-source model aimed at tackling LLM hallucinations. Instead of a monolithic architecture, it uses 3 specialized pathways for per-token routing. The team reports it catches out-of-distribution gibberish with a 0.93 AUROC and cleanly refuses unknown queries. A ternary version is also in active development to bridge the performance gap with standard float models. The project is fully open-source on GitHub and HuggingFace.

原文链接 →

「模型」频道最新

更多「模型」频道 AI 资讯 →