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 2 / 22 页

51. CVPR 2026REVISOR Beyond Textual Reflection Towards Multimodal Introspective Reasoning in Long-Form
52. CVPR 2026S2-MLLM Boosting Spatial Reasoning Capability of MLLMs for 3D Visual
53. CVPR 2026Saliency-R1 Enforcing Interpretable and Faithful Vision-language Reasoning via Saliency-map Alignment
54. CVPR 2026See Less See Right Bi-directional Perceptual Shaping For Multimodal Reasoning
55. CVPR 2026See What We Cannot See A Geo-guided Reasoning Benchmark for
56. CVPR 2026Seeing Clearly Reasoning Confidently Plug-and-Play Remedies for Vision Language Model
57. CVPR 2026Solving a Nonlinear Blind Inverse Problem for Tagged MRI with
58. CVPR 2026Solving Minimal Problems Without Matrix Inversion Using FFT-Based Interpolation
59. CVPR 2026SpaceMind Camera-Guided Modality Fusion for Spatial Reasoning in Vision-Language Models
60. CVPR 2026SpaceTools Tool-Augmented Spatial Reasoning via Double Interactive RL
61. CVPR 2026Sparse-LaViDa Sparse Multimodal Discrete Diffusion Language Models
62. CVPR 2026SpatiaLQA A Benchmark for Evaluating Spatial Logical Reasoning in Vision-Language
63. CVPR 2026SpatialStack Layered Geometry-Language Fusion for 3D VLM Spatial Reasoning
64. CVPR 2026Stable and Efficient Single-Rollout RL for Multimodal Reasoning
65. CVPR 2026StaR-KVQA Structured Reasoning Traces for Implicit-Knowledge Visual Question Answering
66. CVPR 2026SurgCoT Advancing Spatiotemporal Reasoning in Surgical Videos through a Chain-of-Thought
67. CVPR 2026TableMix Enhancing Multimodal Table Reasoning in MLLMs from a Data-Centric
68. CVPR 2026TerraScope Pixel-Grounded Visual Reasoning for Earth Observation
69. CVPR 2026Think 360deg Beyond Depth Evaluating the Width-centric Reasoning Capability of
70. CVPR 2026Think Visually Reason Textually Vision-Language Synergy in Abstract Reasoning
71. CVPR 2026Think with 3D Geometric Imagination Grounded Spatial Reasoning from Limited
72. CVPR 2026Thinking Diffusion Penalize and Guide Visual-Grounded Reasoning in Diffusion Multimodal
73. CVPR 2026Towards Efficient Medical Reasoning with Minimal Fine-Tuning Data
74. CVPR 2026Towards Reasoning-Preserving Unlearning in Multimodal Large Language Models
75. CVPR 2026UniMERNet A Universal Network for Real-World Mathematical Expression Recognition
76. CVPR 2026Unlocking Token Rewards via Training-Free Reward Attribution
77. CVPR 2026VideoAuto-R1 Video Auto Reasoning via Thinking Once Answering Twice
78. CVPR 2026ViRC Enhancing Visual Interleaved Mathematical CoT with Reason Chunking
79. CVPR 2026VIRO Robust and Efficient Neuro-Symbolic Reasoning with Verification for Referring
80. CVPR 2026VisRes Bench On Evaluating the Visual Reasoning Capabilities of VLMs
81. CVPR 2026When Visualizing is the First Step to Reasoning MIRA a
82. ICLR 2026$\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Textual Space
83. ICLR 2026$\textbf{Re}^{2}$: Unlocking LLM Reasoning via Reinforcement Learning with Re-solving
84. ICLR 2026A State-Transition Framework for Efficient LLM Reasoning
85. ICLR 2026AceReason-Nemotron 1.1: Advancing Math and Code Reasoning through SFT and RL Synergy
86. ICLR 2026Analysis of approximate linear programming solution to Markov decision problem with log barrier function
87. ICLR 2026Are Reasoning LLMs Robust to Interventions on their Chain-of-Thought?
88. ICLR 2026ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning
89. ICLR 2026Asymmetric Proximal Policy Optimization: mini-critics boost LLM reasoning
90. ICLR 2026AutoGPS: Automated Geometry Problem Solving via Multimodal Formalization and Deductive Reasoning
91. ICLR 2026Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty
92. ICLR 2026Beyond Markovian: Reflective Exploration via Bayes-Adaptive RL for LLM Reasoning
93. ICLR 2026Beyond Pass@ 1: Self-Play with Variational Problem Synthesis Sustains RLVR
94. ICLR 2026BeyondBench: Benchmark-Free Evaluation of Reasoning in Language Models
95. ICLR 2026Bird's-eye-view Informed Reasoning Driver
96. ICLR 2026Bottlenecked Transformers: Periodic KV Cache Consolidation for Generalised Reasoning
97. ICLR 2026Buffer Matters: Unleashing the Power of Off-Policy Reinforcement Learning in Large Language Model Reasoning
98. ICLR 2026C-Voting: Confidence-Based Test-Time Voting without Explicit Energy Functions
99. ICLR 2026Characterizing and Mitigating Reasoning Drift in Large Language Models
100. ICLR 2026CircuitSense: A Hierarchical Circuit System Benchmark Bridging Visual Comprehension and Symbolic Reasoning in Engineering Design Process

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。