数学推理论文趋势 2023–2026

主要结论

• 数学推理论文从 2023 年的 57 篇增长到 2026 年的 577 篇，三年间扩大到约 10 倍。
• NeurIPS 是发数学推理论文最多的会议，2023–2026 年共 469 篇，占四个会议总量的 43%。
• 交叉方向里涨得最快的是强化学习：同时属于两个主题的论文从 2023 年的 3 篇增至 2026 年的 140 篇。
• 在「神经符号」的 5 个主题里，它的 2026 年论文数排第 1；同组增长最快的可微推理是 2.5 倍，数学推理是 10 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026A mathematical theory of balancing relational generalization and memorization
• ICML 2026ASyMOB: Algebraic Symbolic Mathematical Operations Benchmark
• ICLR 2026AutoGPS: Automated Geometry Problem Solving via Multimodal Formalization and Deductive Reasoning
• CVPR 2026Affine Perspective-Three-Point Problem
• NeurIPS 2025Are Language Models Efficient Reasoners? A Perspective from Logic Programming
• ICML 2025Autoformulation of Mathematical Optimization Models Using LLMs
• ICLR 2025Advancing Mathematical Reasoning in Language Models: The Impact of Problem-Solving Data, Data Synthesis Methods, and Training Stages
• CVPR 2025Bridging Viewpoint Gaps: Geometric Reasoning Boosts Semantic Correspondence
• NeurIPS 2024A Careful Examination of Large Language Model Performance on Grade School Arithmetic
• ICML 2024Accelerating Legacy Numerical Solvers by Non-intrusive Gradient-based Meta-solving
• ICLR 2024Don't Trust: Verify -- Grounding LLM Quantitative Reasoning with Autoformalization
• CVPR 2024E-GPS: Explainable Geometry Problem Solving via Top-Down Solver and Bottom-Up Generator
• NeurIPS 2023Evaluating and Improving Tool-Augmented Computation-Intensive Math Reasoning
• ICML 2023Automated Search for Conjectures on Mathematical Constants using Analysis of Integer Sequences
• ICLR 2023Differentiable Mathematical Programming for Object-Centric Representation Learning
• CVPR 2023Abstract Visual Reasoning: An Algebraic Approach for Solving Raven's Progressive Matrices
• NeurIPS 2026A Mechanistic Analysis of Looped Reasoning Language Models
• ICML 2026Beyond Theorem Proving: Formulation, Framework and Benchmark for Formal Problem-Solving
• ICLR 2026CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solving
• CVPR 2026Hilbert-Geo Solving Solid Geometric Problems by Neural-Symbolic Reasoning

全部数学推理论文

共 175 篇，第 1 / 4 页

1. ICLR 2026$\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Textual Space
2. ICLR 2026$\textbf{Re}^{2}$: Unlocking LLM Reasoning via Reinforcement Learning with Re-solving
3. ICLR 2026A State-Transition Framework for Efficient LLM Reasoning
4. ICLR 2026AceReason-Nemotron 1.1: Advancing Math and Code Reasoning through SFT and RL Synergy
5. ICLR 2026Analysis of approximate linear programming solution to Markov decision problem with log barrier function
6. ICLR 2026Are Reasoning LLMs Robust to Interventions on their Chain-of-Thought?
7. ICLR 2026ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning
8. ICLR 2026Asymmetric Proximal Policy Optimization: mini-critics boost LLM reasoning
9. ICLR 2026AutoGPS: Automated Geometry Problem Solving via Multimodal Formalization and Deductive Reasoning
10. ICLR 2026Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty
11. ICLR 2026Beyond Markovian: Reflective Exploration via Bayes-Adaptive RL for LLM Reasoning
12. ICLR 2026Beyond Pass@ 1: Self-Play with Variational Problem Synthesis Sustains RLVR
13. ICLR 2026BeyondBench: Benchmark-Free Evaluation of Reasoning in Language Models
14. ICLR 2026Bird's-eye-view Informed Reasoning Driver
15. ICLR 2026Bottlenecked Transformers: Periodic KV Cache Consolidation for Generalised Reasoning
16. ICLR 2026Buffer Matters: Unleashing the Power of Off-Policy Reinforcement Learning in Large Language Model Reasoning
17. ICLR 2026C-Voting: Confidence-Based Test-Time Voting without Explicit Energy Functions
18. ICLR 2026Characterizing and Mitigating Reasoning Drift in Large Language Models
19. ICLR 2026CircuitSense: A Hierarchical Circuit System Benchmark Bridging Visual Comprehension and Symbolic Reasoning in Engineering Design Process
20. ICLR 2026CMT-Benchmark: A Benchmark for Condensed Matter Theory Built by Expert Researchers
21. ICLR 2026CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solving
22. ICLR 2026Compositional-ARC: Assessing Systematic Generalization in Abstract Spatial Reasoning
23. ICLR 2026Conditional Advantage Estimation for Reinforcement Learning in Large Reasoning Models
24. ICLR 2026Continuous Chain of Thought Enables Parallel Exploration and Reasoning
25. ICLR 2026CORE: Concept-Oriented Reinforcement for Bridging the Definition–Application Gap in Mathematical Reasoning
26. ICLR 2026CoT-Evo: Evolutionary Distillation of Chain-of-Thought for Scientific Reasoning
27. ICLR 2026Curriculum Reinforcement Learning from Easy to Hard Tasks Improves LLM Reasoning
28. ICLR 2026CyclicReflex: Improving Reasoning Models via Cyclical Reflection Token Scheduling
29. ICLR 2026DAG-Math: Graph-Guided Mathematical Reasoning in LLMs
30. ICLR 2026Deep Think with Confidence
31. ICLR 2026DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning
32. ICLR 2026DIVA-GRPO: Enhancing Multimodal Reasoning through Difficulty-Adaptive Variant Advantage
33. ICLR 2026Diversity-Enhanced Reasoning for Subjective Questions
34. ICLR 2026Divide and Abstract: Autoformalization via Decomposition and Abstraction Learning
35. ICLR 2026DRAGON: Guard LLM Unlearning in Context via Negative Detection and Reasoning
36. ICLR 2026DRIFT: Decompose, Retrieve, Illustrate, then Formalize Theorems
37. ICLR 2026DRPO: Efficient Reasoning via Decoupled Reward Policy Optimization
38. ICLR 2026Dynamic Early Exit in Reasoning Models
39. ICLR 2026Echoes as Anchors: Probabilistic Costs and Attention Refocusing in LLM Reasoning
40. ICLR 2026Efficient Reasoning with Balanced Thinking
41. ICLR 2026Emergent Hierarchical Reasoning in LLMs through Reinforcement Learning
42. ICLR 2026Energy-Based Transformers are Scalable Learners and Thinkers
43. ICLR 2026Equilibrium Language Models
44. ICLR 2026Evaluating Cross-Modal Reasoning Ability and Problem Charactaristics with Multimodal Item Response Theory
45. ICLR 2026Evaluating Language Models' Evaluations of Games
46. ICLR 2026EvolProver: Advancing Automated theorem proving by Evolving Formalized Problems via Symmetry and Difficulty
47. ICLR 2026ExGRPO: Learning to Reason from Prior Successes
48. ICLR 2026Expanding Reasoning Potential in Foundation Model by Learning Diverse Chains of Thought Patterns
49. ICLR 2026Explain in Your Own Words: Improving Reasoning via Token-Selective Dual Knowledge Distillation
50. ICLR 2026FaithCoT-Bench: Benchmarking Instance-Level Faithfulness of Chain-of-Thought Reasoning

常见问题

2023–2026 年四大会议一共有多少篇数学推理论文？

共 1,085 篇，按年份分别是 2023 年 57 篇、2024 年 137 篇、2025 年 314 篇、2026 年 577 篇。

NeurIPS 2026 有多少篇数学推理论文？

NeurIPS 2026 有 208 篇数学推理论文，上一届是 190 篇。

ICLR 2026 有多少篇数学推理论文？

ICLR 2026 有 175 篇数学推理论文，上一届是 61 篇。

ICML 2026 有多少篇数学推理论文？

ICML 2026 有 113 篇数学推理论文，上一届是 45 篇。

CVPR 2026 有多少篇数学推理论文？

CVPR 2026 有 81 篇数学推理论文，上一届是 18 篇。

哪个会议的数学推理论文最多？

NeurIPS，2023–2026 年共 469 篇。