Proof Artifact Co-training for Theorem Proving with Language Models
Abstract
Domain fit: AI-adjacent · Paper appears method- or tooling-adjacent to AI workflows with partial ecosystem coverage.
Labeled data for imitation learning of theorem proving in large libraries of formalized mathematics is scarce as such libraries require years of concentrated effort by human specialists to be built. This is particularly challenging when applying large Transformer language models to tactic prediction, because the scaling of performance with respect to model size is quickly disrupted in the data-scarce, easily-overfitted regime. We propose PACT ({\bf P}roof {\bf A}rtifact {\bf C}o-{\bf T}raining), a general methodology for extracting abundant self-supervised data from kernel-level proof terms for co-training alongside the usual tactic prediction objective. We apply this methodology to Lean, an interactive proof assistant which hosts some of the most sophisticated formalized mathematics to date. We instrument Lean with a neural theorem prover driven by a Transformer language model and show that PACT improves theorem proving success rate on a held-out suite of test theorems from 32\% to 48\%.
Results and benchmarks
Labeled data for imitation learning of theorem proving in large libraries of formalized mathematics is scarce as such libraries require years of concentrated effort by human specialists to be built.
Benchmark evidence is limited
Evidence graph: 2 refs, 1 links.
Utility signals: depth 65/100, grounding 58/100, status medium.
Implementation
No direct implementation yet
Maintained implementation evidence is not confirmed for this paper yet.
Use the implementation status and reproduction sections for the current action plan.
No verified maintained repo yet
There is no verified maintained implementation yet. Use this baseline plan to decide whether to prototype now or defer.
- No direct maintained implementation was found. Use the paper PDF and citation graph to design a baseline reproduction.
- Start from related paper: Tactic-based theorem proving and knowledge-based forward chaining: An experiment with Nuprl and Ontic.
- Start from this likely method family: Transformer.
Time to first repro: a few days
Recommendation evidence is currently too limited for a maintained-repo choice. Use Implementation Status and Reproduction Path for a practical baseline plan.
- Estimate is based on paper-only reproduction flow
Reproduction readiness
No repo
No verified implementation available
- No maintained repository has been identified for this paper. Check adjacent implementations or HF artifacts below.
Hardware requirements
- Expect multi-day setup/compute for meaningful reproduction based on current guidance.
Validation caveat
Framework baselines
- Hugging Face Transformers training guide
Modern transformer training baseline.
- PyTorch nn.Transformer docs
Reference transformer building block implementation.
Hugging Face artifacts
No trustworthy direct or curated related Hugging Face artifacts were found yet. Use targeted searches to quickly locate candidate models, datasets, and demos.
Tip: start with models, then check datasets and spaces if you need evaluation data or demos.
Research context
6
Citations
57
References
Tasks
Automated theorem proving, Proof assistant, Computer science, Benchmarking, Suite, Proof of concept, Training set, Theoretical computer science
Methods
Transformer
Domains
Artificial intelligence, Machine learning, Mathematics
Related papers
- Tactic-based theorem proving and knowledge-based forward chaining: An experiment with Nuprl and OnticSearch on Paper2Code
1992 · Semantic similarity
- Analysis and Transformation of Proof ProceduresSearch on Paper2Code
1994 · Semantic similarity
- A Deep Reinforcement Learning Approach to First-Order Logic Theorem ProvingSearch on Paper2Code
2021 · Semantic similarity
- Portfolio theorem proving and prover runtime prediction for geometrySearch on Paper2Code
2018 · Semantic similarity
- Theorem Proving in LeanSearch on Paper2Code
2018 · Semantic similarity
- Generative Language Modeling for Automated Theorem ProvingSearch on Paper2Code
2020 · Semantic similarity
Open this paper in HFEPX to review benchmark signals, evaluation modes, and human-feedback protocol context.
Open in HFEPXJump to Paper2Code search queries derived from this paper's research context.