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 208 篇，第 1 / 5 页

1. NeurIPS 20262D Spatial Reasoning with Adaptive Neural Cellular Automata
2. NeurIPS 2026A mathematical theory of balancing relational generalization and memorization
3. NeurIPS 2026A Matter of Interest: Understanding Interestingness Judgments of Math Problems in Humans and Language Models
4. NeurIPS 2026A Mechanistic Analysis of Looped Reasoning Language Models
5. NeurIPS 2026A Model of Diverse Sampling from Language Models
6. NeurIPS 2026A Theory of Online Learning with Autoregressive Chain-of-Thought Reasoning
7. NeurIPS 2026A$^3$Bench: A Benchmark for Compositional Reasoning over Aggressive Interactions in Videos
8. NeurIPS 2026Active Probabilistic Reasoning in Humans and LLMs
9. NeurIPS 2026AdaKVQ: Adaptive Mixed-Precision KV Cache Quantization For Efficient Reasoning Models
10. NeurIPS 2026Agentic Geometry Problem Solving via Human-like Parallel Bidirectional Reasoning
11. NeurIPS 2026Agentic Multi-Turn Reasoning: A Fairness Approach
12. NeurIPS 2026Allocentric Perceiver: Disentangling Allocentric Reasoning from Egocentric Visual Priors via Frame Instantiation
13. NeurIPS 2026An Enigma of Artificial Reason: Investigating the Production-Evaluation Gap in Large Reasoning Models
14. NeurIPS 2026Anchoring Reasoning Distillation via Syntactic Constraints
15. NeurIPS 2026Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information
16. NeurIPS 2026Argument Graph Uncertainty: Quantifying Uncertainty from the Logical Structure of Reasoning Chains
17. NeurIPS 2026ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both
18. NeurIPS 2026Auxiliary Clues Aware’s Geometry Problem Solving
19. NeurIPS 2026AVSD: Adaptive-View Self-Distillation by Balancing Consensus and Teacher-Specific Privileged Signals
20. NeurIPS 2026Bellman Contraction under MMD: A Unified Framework
21. NeurIPS 2026Benchmarking Multimodal Mathematical Reasoning with Explicit Visual Dependency
22. NeurIPS 2026Beyond Global Alignment: Structured Compositional Reasoning for Vision-Language Models
23. NeurIPS 2026Beyond Real or Fake: A Dual-Channel Authenticity and Reasoning Protocol for Photographic Assessment
24. NeurIPS 2026Beyond What Seems Necessary: Hidden Gains from Scaling Training-Time Reasoning Length under Outcome Supervision
25. NeurIPS 2026Bias-Variance Optimized Preference Optimization for Large Reasoning Models
26. NeurIPS 2026BoardGameArena: A Multi-Dimensional Benchmark for Strategic Reasoning of LLMs in Board Games
27. NeurIPS 2026Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
28. NeurIPS 2026Boosting LLM Reasoning via Human-Inspired Reward Shaping
29. NeurIPS 2026Boule or Baguette? A Study on Task Topology, Length Generalization, and the Benefit of Reasoning Trace
30. NeurIPS 2026Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning
31. NeurIPS 2026Can Linguistic Reasoning Vectors Enhance Multimodal Reasoning Ability?
32. NeurIPS 2026Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations
33. NeurIPS 2026CasePlay: Self-Play Reinforcement Learning from Case Reports for Medical Reasoning
34. NeurIPS 2026CAVE: A Structured Credit Assignment Approach for Fragmented Visual Evidence Reasoning
35. NeurIPS 2026CellularS2-Bench: A Staged, Evidence-Grounded Benchmark for Cellular Network Security Reasoning
36. NeurIPS 2026ChainSpace: A Chained-Reasoning Paradigm for Spatial Intelligence
37. NeurIPS 2026CineMME: Benchmarking Fine-Grained Perception and Plot Reasoning in Multimodal Large Language Models
38. NeurIPS 2026CircuitSeer: Mining High-Quality Data by Probing Mathematical Reasoning Circuits in LLMs
39. NeurIPS 2026Collaborative Reasoning Distillation via Cross-Feedback and Coherent Curation
40. NeurIPS 2026Commit Then Explore: Reasoning Models Diversify, Not Converge
41. NeurIPS 2026Compositional Generalization Certificates via the Van Kampen Theorem
42. NeurIPS 2026Concise Reasoning Through the Lens of Lagrangian Optimization
43. NeurIPS 2026Confidence-Calibrated Inference Expansion for Evaluator-Guided Test-Time Reasoning
44. NeurIPS 2026Consolidating Reasoning with Test-Time Learning
45. NeurIPS 2026Contractive Restoring Flows: Robust Reasoning Distillation via Orbital Stability
46. NeurIPS 2026Contrastive Thinking Decoding: Steering Answer Generation in Reasoning Models
47. NeurIPS 2026Convex Compositional Reasoning Models
48. NeurIPS 2026CoRE-RL: Co-evolving Reasoning Trajectories and Evidence Subgraphs with Learned Evidence Projection
49. NeurIPS 2026Correct but Unselectable: The Hidden Interface Tax in Multi-Candidate Reasoning
50. NeurIPS 2026Corrective Diffusion Language Models

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。