Reproduce “A Neurosymbolic Approach to Natural Language Formalization and Verification” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chenyang An, Sam Bayless, Stefano Buliani, Darion Cassel, Byron Cook, Duncan Clough, Rémi Delmas, Nafi Diallo and 22 more
ARc checks whether a chatbot answer follows a formalized policy and stated conditions. It uses language models to translate the policy and answer into logic, then asks a solver to classify the result. On ConditionalQA-logic, three translations with a 3/3 confidence threshold achieved 99.4% soundness and 14.9% recall; results depend on policy formalization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.