┌─────────────┐ ┌──────────────┐ ┌────────────────┐ ┌────� ...
This repository implements an Isabelle/HOL theorem prover guided by Large Language Models (LLMs). It integrates Isabelle’s proof engine with modern LLMs (via Ollama, Gemini CLI, etc).
They struggle to adapt to strategies outside their training distribution. In contrast, our ‘Brain-Torso' architecture maintains a robust Average Win Rate (AWR) of 70.9%. The Strategic Brain utilizes ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
