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 314 篇，第 2 / 7 页

51. ICLR 2025Mitigating Modality Prior-Induced Hallucinations in Multimodal Large Language Models via Deciphering Attention Causality
52. ICLR 2025MLPs Learn In-Context on Regression and Classification Tasks
53. ICLR 2025MR-GSM8K: A Meta-Reasoning Benchmark for Large Language Model Evaluation
54. ICLR 2025Multiagent Finetuning: Self Improvement with Diverse Reasoning Chains
55. ICLR 2025Mutual Reasoning Makes Smaller LLMs Stronger Problem-Solver
56. ICLR 2025Number Cookbook: Number Understanding of Language Models and How to Improve It
57. ICLR 2025Omni-MATH: A Universal Olympiad Level Mathematic Benchmark for Large Language Models
58. ICLR 2025OpenMathInstruct-2: Accelerating AI for Math with Massive Open-Source Instruction Data
59. ICLR 2025Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process
60. ICLR 2025Physics of Language Models: Part 2.2, How to Learn From Mistakes on Grade-School Math Problems
61. ICLR 2025Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models
62. ICLR 2025Process Reward Model with Q-value Rankings
63. ICLR 2025Progress or Regress? Self-Improvement Reversal in Post-training
64. ICLR 2025Proving Olympiad Inequalities by Synergizing LLMs and Symbolic Reasoning
65. ICLR 2025Reasoning Elicitation in Language Models via Counterfactual Feedback
66. ICLR 2025Reasoning with Latent Thoughts: On the Power of Looped Transformers
67. ICLR 2025ReGenesis: LLMs can Grow into Reasoning Generalists via Self-Improvement
68. ICLR 2025SBSC: Step-by-Step Coding for Improving Mathematical Olympiad Performance
69. ICLR 2025Scaling LLM Test-Time Compute Optimally Can be More Effective than Scaling Parameters for Reasoning
70. ICLR 2025Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning
71. ICLR 2025Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling
72. ICLR 2025Systematic Relational Reasoning With Epistemic Graph Neural Networks
73. ICLR 2025ThinkBot: Embodied Instruction Following with Thought Chain Reasoning
74. ICLR 2025To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning
75. ICLR 2025Token-Supervised Value Models for Enhancing Mathematical Problem-Solving Capabilities of Large Language Models
76. ICLR 2025Transformer Encoder Satisfiability: Complexity and Impact on Formal Reasoning
77. ICLR 2025TypedThinker: Diversify Large Language Model Reasoning with Typed Thinking
78. ICLR 2025UGMathBench: A Diverse and Dynamic Benchmark for Undergraduate-Level Mathematical Reasoning with Large Language Models
79. ICLR 2025WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instruct
80. ICML 2025A Mathematical Framework for AI-Human Integration in Work
81. ICML 2025A Reasoning-Based Approach to Cryptic Crossword Clue Solving
82. ICML 2025A Sub-Problem Quantum Alternating Operator Ansatz for Correlation Clustering
83. ICML 2025AdaptiveStep: Automatically Dividing Reasoning Step through Model Confidence
84. ICML 2025Attention-Level Speculation
85. ICML 2025Autoformulation of Mathematical Optimization Models Using LLMs
86. ICML 2025Beyond Task-Specific Reasoning: A Unified Conditional Generative Framework for Abstract Visual Reasoning
87. ICML 2025Blink of an eye: a simple theory for feature localization in generative models
88. ICML 2025Can Transformers Reason Logically? A Study in SAT Solving
89. ICML 2025Discrete Neural Algorithmic Reasoning
90. ICML 2025Do NOT Think That Much for 2+3=? On the Overthinking of Long Reasoning Models
91. ICML 2025Fleet of Agents: Coordinated Problem Solving with Large Language Models
92. ICML 2025Forest-of-Thought: Scaling Test-Time Compute for Enhancing LLM Reasoning
93. ICML 2025Free Process Rewards without Process Labels
94. ICML 2025From Language Models over Tokens to Language Models over Characters
95. ICML 2025Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models
96. ICML 2025GSM-$\infty$: How Do your LLMs Behave over Infinitely Increasing Reasoning Complexity and Context Length?
97. ICML 2025HyperTree Planning: Enhancing LLM Reasoning via Hierarchical Thinking
98. ICML 2025Improving Rationality in the Reasoning Process of Language Models through Self-playing Game
99. ICML 2025Leveraging Online Olympiad-Level Math Problems for LLMs Training and Contamination-Resistant Evaluation
100. ICML 2025Lower Bounds for Chain-of-Thought Reasoning in Hard-Attention Transformers

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。