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 137 篇，第 2 / 3 页

51. ICML 2024Attention Meets Post-hoc Interpretability: A Mathematical Perspective
52. ICML 2024Auto-Regressive Next-Token Predictors are Universal Learners
53. ICML 2024Case-Based or Rule-Based: How Do Transformers Do the Math?
54. ICML 2024ConTextual: Evaluating Context-Sensitive Text-Rich Visual Reasoning in Large Multimodal Models
55. ICML 2024ContPhy: Continuum Physical Concept Learning and Reasoning from Videos
56. ICML 2024Counterfactual Reasoning for Multi-Label Image Classification via Patching-Based Training
57. ICML 2024CRUXEval: A Benchmark for Code Reasoning, Understanding and Execution
58. ICML 2024Do Language Models Exhibit the Same Cognitive Biases in Problem Solving as Human Learners?
59. ICML 2024Enforcing Constraints in RNA Secondary Structure Predictions: A Post-Processing Framework Based on the Assignment Problem
60. ICML 2024GeoReasoner: Geo-localization with Reasoning in Street Views using a Large Vision-Language Model
61. ICML 2024GLoRe: When, Where, and How to Improve LLM Reasoning via Global and Local Refinements
62. ICML 2024Graph2Tac: Online Representation Learning of Formal Math Concepts
63. ICML 2024How Smooth Is Attention?
64. ICML 2024Improving Factuality and Reasoning in Language Models through Multiagent Debate
65. ICML 2024INViT: A Generalizable Routing Problem Solver with Invariant Nested View Transformer
66. ICML 2024Language Agent Tree Search Unifies Reasoning, Acting, and Planning in Language Models
67. ICML 2024Language Models with Conformal Factuality Guarantees
68. ICML 2024Learning Iterative Reasoning through Energy Diffusion
69. ICML 2024Learning Solution-Aware Transformers for Efficiently Solving Quadratic Assignment Problem
70. ICML 2024MathScale: Scaling Instruction Tuning for Mathematical Reasoning
71. ICML 2024MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities
72. ICML 2024Neural operators meet conjugate gradients: The FCG-NO method for efficient PDE solving
73. ICML 2024On a Combinatorial Problem Arising in Machine Teaching
74. ICML 2024On the Hardness of Probabilistic Neurosymbolic Learning
75. ICML 2024Partial Optimality in the Linear Ordering Problem
76. ICML 2024Position: Rethinking Post-Hoc Search-Based Neural Approaches for Solving Large-Scale Traveling Salesman Problems
77. ICML 2024Premise Order Matters in Reasoning with Large Language Models
78. ICML 2024Rethinking Transformers in Solving POMDPs
79. ICML 2024SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models
80. ICML 2024Slot Abstractors: Toward Scalable Abstract Visual Reasoning
81. ICML 2024Solving Poisson Equations using Neural Walk-on-Spheres
82. ICML 2024Toward Adaptive Reasoning in Large Language Models with Thought Rollback
83. ICML 2024Training Large Language Models for Reasoning through Reverse Curriculum Reinforcement Learning
84. ICML 2024TroVE: Inducing Verifiable and Efficient Toolboxes for Solving Programmatic Tasks
85. ICML 2024Video-of-Thought: Step-by-Step Video Reasoning from Perception to Cognition
86. NeurIPS 2024A Careful Examination of Large Language Model Performance on Grade School Arithmetic
87. NeurIPS 2024A Neuro-Symbolic Benchmark Suite for Concept Quality and Reasoning Shortcuts
88. NeurIPS 2024Abductive Reasoning in Logical Credal Networks
89. NeurIPS 2024AlphaMath Almost Zero: Process Supervision without Process
90. NeurIPS 2024Autoformalize Mathematical Statements by Symbolic Equivalence and Semantic Consistency
91. NeurIPS 2024AutoPSV: Automated Process-Supervised Verifier
92. NeurIPS 2024Buffer of Thoughts: Thought-Augmented Reasoning with Large Language Models
93. NeurIPS 2024Calibrating Reasoning in Language Models with Internal Consistency
94. NeurIPS 2024Can Language Models Learn to Skip Steps?
95. NeurIPS 2024Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?
96. NeurIPS 2024Can LLMs Learn by Teaching for Better Reasoning? A Preliminary Study
97. NeurIPS 2024Chain-of-Thought Reasoning Without Prompting
98. NeurIPS 2024ClevrSkills: Compositional Language And Visual Reasoning in Robotics
99. NeurIPS 2024DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving
100. NeurIPS 2024Decompose, Analyze and Rethink: Solving Intricate Problems with Human-like Reasoning Cycle

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。