Get Started
Home
Topics
Search
Library
Research questionHow can autoformalization preserve diverse faithful statements that improve prover search under a fixed budget?A single formal translation can hide other faithful formulations, while syntactic differences among equivalent statements can change how a prover searches. Correctness-only, single-output evaluation therefore misses effects that matter for downstream proving.
AI
Code Generation & Program Synthesis
Neural and Evolutionary Computing
Reasoning
Latest papersRecent research connected to this question, newest first.FormalEvolve: Neuro-Symbolic Evolutionary Search for Diverse AutoformalizationThe evidence concerns FormalEvolve’s LLM-driven autoformalization search, which maintains compilation-feasible candidates and uses a fixed LLM semantic judge to produce deduplicated, semantically accepted repertoires. Results use 100 generator calls on CombiBench and ProofNet and a fixed prover budget of 64; stronger seed and repair models are also tested. The evidence is limited to these settings and reported budgets.research paper · Sep 2, 2026
Related questions
How can automated Lean 4 proof search reuse promising partial proofs without later revisions degrading them?How can we detect when a compiling Lean formalization changes the intended natural-language meaning?Can any finite formal system autonomously derive every theorem within its expressive scope?How can LLM agents discover auxiliary constructions that support formally verified geometry proofs?