LeanPRM auto-labels proof steps with Lean, enabling verifier training without humans
Rodney Lafuente-Mercado · hf · 2026-10-17
LeanPRM (UT Austin/UCLA/NYU) uses the Lean proof assistant to generate step-level correctness labels for natural-language math proofs: a model writes a proof paired with a formal version, and Lean grades each step — replacing costly human annotation and enabling scalable training of process reward verifiers for long AI-generated proofs.
More from Research
- LeanGraphSearch mines Mathlib dependency graphs, lifting proof success to 81% — kaiwei_chang · 2026-10-17
- Perplexity decider v1.2 lands #3 on Decision Index as longer training erodes backbone skills — antoine_chaffin · 2026-10-17
- Perplexity traces decider v1.2 regressions to longer training, mitigates via model souping — antoine_chaffin · 2026-10-17
- Packing multiple questions into one forward pass saves compute but adds instability — antoine_chaffin · 2026-10-17
- pplx-decider packs multiple questions per state into one bidirectional forward pass — antoine_chaffin · 2026-10-17
- Dual-mask fine-tuning recovers cacheability with only -0.4 DI vs full bidirectional — antoine_chaffin · 2026-10-17