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 1,085 篇，第 22 / 22 页

1051. ICLR 2023Protein Representation Learning via Knowledge Enhanced Primary Structure Reasoning
1052. ICLR 2023Scenario-based Question Answering with Interacting Contextual Properties
1053. ICLR 2023Self-Consistency Improves Chain of Thought Reasoning in Language Models
1054. ICLR 2023Solving Constrained Variational Inequalities via a First-order Interior Point-based Method
1055. ICLR 2023UniKGQA: Unified Retrieval and Reasoning for Solving Multi-hop Question Answering Over Knowledge Graph
1056. ICML 2023Alternating Local Enumeration (TnALE): Solving Tensor Network Structure Search with Fewer Evaluations
1057. ICML 2023Automated Search for Conjectures on Mathematical Constants using Analysis of Integer Sequences
1058. ICML 2023Generalization on the Unseen, Logic Reasoning and Degree Curriculum
1059. ICML 2023GibbsDDRM: A Partially Collapsed Gibbs Sampler for Solving Blind Inverse Problems with Denoising Diffusion Restoration
1060. ICML 2023Infusing Lattice Symmetry Priors in Attention Mechanisms for Sample-Efficient Abstract Geometric Reasoning
1061. ICML 2023Neural Inverse Operators for Solving PDE Inverse Problems
1062. ICML 2023Neural Prediction Errors enable Analogical Visual Reasoning in Human Standard Intelligence Tests
1063. ICML 2023Specializing Smaller Language Models towards Multi-Step Reasoning
1064. ICML 2023The Catalog Problem: Clustering and Ordering Variable-Sized Sets
1065. ICML 2023The Monge Gap: A Regularizer to Learn All Transport Maps
1066. ICML 2023Towards Constituting Mathematical Structures for Learning to Optimize
1067. NeurIPS 2023Disentangled Counterfactual Learning for Physical Audiovisual Commonsense Reasoning
1068. NeurIPS 2023Evaluating and Improving Tool-Augmented Computation-Intensive Math Reasoning
1069. NeurIPS 2023Fitting trees to $\ell_1$-hyperbolic distances
1070. NeurIPS 2023Globally solving the Gromov-Wasserstein problem for point clouds in low dimensional Euclidean spaces
1071. NeurIPS 2023Grounding Neural Inference with Satisfiability Modulo Theories
1072. NeurIPS 2023How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model
1073. NeurIPS 2023Interactive Visual Reasoning under Uncertainty
1074. NeurIPS 2023Landscape Surrogate: Learning Decision Losses for Mathematical Optimization Under Partial Information
1075. NeurIPS 2023Learning Reliable Logical Rules with SATNet
1076. NeurIPS 2023Mathematical Capabilities of ChatGPT
1077. NeurIPS 2023MathNAS: If Blocks Have a Role in Mathematical Architecture Design
1078. NeurIPS 2023Neural Algorithmic Reasoning Without Intermediate Supervision
1079. NeurIPS 2023Not All Neuro-Symbolic Concepts Are Created Equal: Analysis and Mitigation of Reasoning Shortcuts
1080. NeurIPS 2023Optimal approximation using complex-valued neural networks
1081. NeurIPS 2023Robust Model Reasoning and Fitting via Dual Sparsity Pursuit
1082. NeurIPS 2023SPRING: Studying Papers and Reasoning to play Games
1083. NeurIPS 2023Towards Revealing the Mystery behind Chain of Thought: A Theoretical Perspective
1084. NeurIPS 2023Unsupervised Learning for Solving the Travelling Salesman Problem
1085. NeurIPS 2023Why think step by step? Reasoning emerges from the locality of experience

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。