Skip to main content
Search

# Formal Theorem Proving

5 related articles

Subscribe to updates
  1. arXiv

    Megrez moves textbook topology formalization into an MK set-theory library: 262 files and 2,529 Qed proofs indexed

    For Rocq libraries built over Morse–Kelley set theory, Megrez couples a general coding agent with a rebuildable knowledge graph, a short-burst interactive proving loop and a three-check acceptance pipeline to formalize a textbook chapter by chapter. On the point-set topology corpus MKTopos it covers 12 modules, 262 .v files, 3,894 declarations and 2,529 completed Qed proofs across 102,122 lines of proof script; in retrieval benchmarks, exact-identifier lookup at 100,000 nodes drops from 72.29 ms to 1.23 ms in p95 latency.

  2. aclanthology.org

    Failed Lean proofs stop being thrown away: compiler feedback plus dual-model search lifts average pass rate 12.8 points within a pass@32 budget

    In Lean 4 theorem proving that depends on project context, failed proofs used to be discarded as a final filter result, and independent sampling restarted later attempts nearly from scratch. This search framework draws one candidate each from a general-purpose reasoning model and a Lean-specialized prover, uses compiler errors in a pairwise comparison to pick the current-best proof, accepts a revision only when the comparison judges it better, and resamples when refinement stagnates. On seven real Lean 4 projects from miniCTX-v2, within the pass@32 budget it improves average pass rate by 12.8 percentage points over the two base models' average pass@32 while cutting LLM calls by 21.9%.

  3. aclanthology.org

    Theorem-prover training stops scaling data: 500 boundary-hard tasks mined from verified proofs lift both provers

    Reinforcement learning for formal theorem proving is bottlenecked by scarce hard tasks: existing Lean problems are either already solved by current provers or too uninformative. ProofFrontier starts from verified Lean theorem–proof pairs, keeps each theorem statement unchanged, and hides proof-structural guidance around key local nodes so the prover must rediscover intermediate reasoning; base-prover pass rates then select tasks that are solvable in principle but not easily solved. GRPO training on 500 synthesized tasks improves DeepSeek-Prover-V2-7B and Goedel-Prover-V2-8B on miniF2F, ProofNet and PutnamBench over each base model.

  4. arXiv

    Self-advertised method selection lifts formal-proving hit@5 on Putnam from 84.2% to 95.0%

    The work reframes method selection in formal proving as an applicability problem and introduces self-advertisement: in one batched call, each of 82 Method Contracts proposes a target, an action, and required conditions for the current goal, and vague or unsupported proposals are demoted before ranking; it reaches 95.0% hit@5 on Putnam 2015–2025 versus 84.2% for the strongest reranker, and 91.7% versus 88.3% on IMO ProofBench.

  5. arXiv

    Lean Pool: An AI-Maintained Archive of Formalized Mathematics

    Lean Pool is a repository of formalized mathematics grown, maintained, and optimized by AI agents; the paper reports that as of September 21, 2026 it holds 211 completed projects and 3,228,485 lines of Lean code from 18 commit contributors, and that agent-assisted dependency upgrades, proof compression, and mathematical review keep these independently developed formalizations usable as Lean and Mathlib evolve.

You’re all caught up