SOTAVerified

Automated Theorem Proving

The goal of Automated Theorem Proving is to automatically generate a proof, given a conjecture (the target theorem) and a knowledge base of known facts, all expressed in a formal language. Automated Theorem Proving is useful in a wide range of applications, including the verification and synthesis of software and hardware systems.

Source: Learning to Prove Theorems by Learning to Generate Theorems

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–100 of 288 papers

TitleStatusHype
Chain-of-Reasoning: Towards Unified Mathematical Reasoning in Large Language Models via a Multi-Paradigm Perspective—0
Proof Recommendation System for the HOL4 Theorem Prover—0
HUNYUANPROVER: A Scalable Data Synthesis Framework and Guided Tree Search for Automated Theorem Proving—0
Formal Mathematical Reasoning: A New Frontier in AI—0
Towards Scientific Discovery with Generative AI: Progress, Opportunities, and Challenges—0
WithdrarXiv: A Large-Scale Dataset for Retraction StudyCode0
Improving Multimodal LLMs Ability In Geometry Problem Solving, Reasoning, And Multistep Scoring—0
Formal Theorem Proving by Rewarding LLMs to Decompose Proofs Hierarchically—0
Learning Rules Explaining Interactive Theorem Proving Tactic PredictionCode0
Beyond Autoregression: Fast LLMs via Self-Distillation Through TimeCode1
Pantograph: A Machine-to-Machine Interaction Interface for Advanced Theorem Proving, High Level Reasoning, and Data Extraction in Lean 4Code2
Alchemy: Amplifying Theorem-Proving Capability through Symbolic MutationCode0
InternLM2.5-StepProver: Advancing Automated Theorem Proving via Expert Iteration on Large-Scale LEAN ProblemsCode4
Proof Flow: Preliminary Study on Generative Flow Network Language Model Tuning for Formal Reasoning—0
3D-Prover: Diversity Driven Theorem Proving With Determinantal Point Processes—0
LeanAgent: Lifelong Learning for Formal Theorem ProvingCode2
Mathematical Formalized Problem Solving and Theorem Proving in Different Fields in Lean 4Code0
SubgoalXL: Subgoal-based Expert Learning for Theorem ProvingCode1
DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree SearchCode4
Revealed Invariant Preference—0
miniCTX: Neural Theorem Proving with (Long-)ContextsCode4
Artifical intelligence and inherent mathematical difficulty—0
LEAN-GitHub: Compiling GitHub LEAN repositories for a versatile LEAN proverCode4
PutnamBench: Evaluating Neural Theorem-Provers on the Putnam Mathematical CompetitionCode3
Lean-STaR: Learning to Interleave Thinking and Proving—0
Towards Automated Functional Equation Proving: A Benchmark Dataset and A Domain-Specific In-Context Agent—0
TheoremLlama: Transforming General-Purpose LLMs into Lean4 ExpertsCode1
Learning Formal Mathematics From Intrinsic MotivationCode2
FVEL: Interactive Formal Verification Environment with Large Language Models via Theorem ProvingCode0
Scaling Synthetic Logical Reasoning Datasets with Context-Sensitive Declarative GrammarsCode0
miniCodeProps: a Minimal Benchmark for Proving Code Properties—0
Lean Workbook: A large-scale Lean problem set formalized from natural language math problemsCode4
DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Data—0
Proving Theorems RecursivelyCode1
A Certified Proof Checker for Deep Neural Network Verification in Imandra—0
ATG: Benchmarking Automated Theorem Generation for Generative Language Models—0
Verification and Refinement of Natural Language Explanations through LLM-Symbolic Theorem ProvingCode0
Lean Copilot: Large Language Models as Copilots for Theorem Proving in LeanCode5
A Survey on Deep Learning for Theorem ProvingCode3
Learn from Failure: Fine-Tuning LLMs with Trial-and-Error Data for Intuitionistic Propositional Logic ProvingCode0
Wu's Method can Boost Symbolic AI to Rival Silver Medalists and AlphaGeometry to Outperform Gold Medalists at IMO Geometry—0
Proceedings 12th International Workshop on Theorem proving components for Educational software—0
Don't Trust: Verify -- Grounding LLM Quantitative Reasoning with AutoformalizationCode1
Multi-Task Learning with Multi-Task Optimization—0
LeanReasoner: Boosting Complex Logical Reasoning with LeanCode1
Enhancing Formal Theorem Proving: A Comprehensive Dataset for Training AI Models on Coq Code—0
Learning Guided Automated Reasoning: A Brief Survey—0
BAIT: Benchmarking (Embedding) Architectures for Interactive Theorem-Proving—0
A Categorization of Complexity Classes for Information Retrieval and Synthesis Using Natural Logic—0
On the (In)feasibility of ML Backdoor Detection as an Hypothesis Testing ProblemCode0
Show:102550
← PrevPage 2 of 6Next →
All datasetsminiF2F-testminiF2F-validHolStep (Conditional)HOList benchmarkHolStep (Unconditional)Metamath set.mmminiF2F-curriculumCompCertCoqGym

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Kimina-Prover-Previewcumulative80.74—Unverified
2ProofAugcumulative66—Unverified
3DeepSeek-Prover-V1.5cumulative63.5—Unverified
4Subgoal-XLcumulative56.1—Unverified
5DeepSeek-Provercumulative52—Unverified
6Lyra + GPT-4cumulative47.1—Unverified
7LEGO-Prover ChatGPTcumulative47.1—Unverified
8Decomposing the Enigmacumulative45.5—Unverified
9Evaristecumulative41—Unverified
10Evariste-7dcumulative40.6—Unverified
#ModelMetricClaimedVerifiedStatus
1EvaristePass@6458.6—Unverified
2LEGO-Prover ChatGPTPass@10057—Unverified
3Lyra + GPT-4Pass@10052—Unverified
4Evariste-7dPass@6447.5—Unverified
5GPT-fPass@6447.3—Unverified
6Evariste-1dPass@6446.7—Unverified
7DSP (62B Minerva informal)Pass@10043.9—Unverified
8Lean GPT-fPass@829.3—Unverified
9Lean tidyPass@116.8—Unverified
10Metamath GPT-fPass@82—Unverified
#ModelMetricClaimedVerifiedStatus
1MPNN-DagLSTMClassification Accuracy0.92—Unverified
2FormulaNetClassification Accuracy0.9—Unverified
3FormulaNet-basicClassification Accuracy0.89—Unverified
4Siamese 1D CNN-LSTMClassification Accuracy0.83—Unverified
5Siamese 1D CNNClassification Accuracy0.82—Unverified
#ModelMetricClaimedVerifiedStatus
14-hop GNN, sub-expression sharingPercentage correct49.95—Unverified
2Tactic Dependent LoopPercentage correct38.88—Unverified
3BoW2 (extra -ves)Percentage correct36.55—Unverified
4Deeper Wider WaveNetPercentage correct32.65—Unverified
#ModelMetricClaimedVerifiedStatus
1FormulaNetClassification Accuracy0.9—Unverified
2FormulaNet-basicClassification Accuracy0.89—Unverified
31D CNN-LSTMClassification Accuracy0.83—Unverified
41D CNNClassification Accuracy0.83—Unverified
#ModelMetricClaimedVerifiedStatus
1EvaristePass@3272.4—Unverified
2GPT-fPercentage correct56.2—Unverified
3MetaGen-IL + HolophrasmPercentage correct22.1—Unverified
4HolophrasmPercentage correct14.3—Unverified
#ModelMetricClaimedVerifiedStatus
1Evariste-7dPass@6442.5—Unverified
2Evariste-1dPass@6433.6—Unverified
3EvaristePass@6432.1—Unverified
4GPT-fPass@6430.6—Unverified
#ModelMetricClaimedVerifiedStatus
1Proverbot9001Percentage correct19.36—Unverified
2CoqGym/ASTacticPercentage correct4.99—Unverified
#ModelMetricClaimedVerifiedStatus
1ASTacticPercentage correct12.2—Unverified