ReasonSTL: Bridging Natural Language and Signal Temporal Logic via Tool-Augmented Process-Rewarded Learning
Researchers have developed ReasonSTL, a novel framework designed to translate natural language requirements into Signal Temporal Logic (STL) formulas. This tool-augmented approach utilizes local, open-source language models to perform the translation, addressing concerns about cost and privacy associated with commercial LLM APIs. ReasonSTL breaks down the process into reasoning, tool calls, and formula construction, incorporating process-rewarded training and a new benchmark called STL-Bench. AI
IMPACT Provides a privacy-preserving, low-cost method for generating formal specifications, potentially improving the verification of autonomous and cyber-physical systems.