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 577 篇，第 10 / 12 页

451. NeurIPS 2026How do Small Transformer Models Learn Hard Math Tasks?
452. NeurIPS 2026How does RL Post-training Induce Skill Composition? A Case Study on Countdown
453. NeurIPS 2026Identified-Set Geometry of Distributional Model Extraction under Top-K Censored API Access
454. NeurIPS 2026INFUSER: Influence-Guided Self-Evolution Improves Reasoning
455. NeurIPS 2026Interactive Combinatorial Reinforcement Learning for Knowledge Graph Reasoning
456. NeurIPS 2026Iterative Nonlinear Computation Underlying Abstract Reasoning
457. NeurIPS 2026Language Models Need Sleep
458. NeurIPS 2026Latent Spatial Reasoning: Building Innate 3D Awareness via Latent-Space Distillation
459. NeurIPS 2026LatentUM: Unleashing the Potential of Interleaved Cross-Modal Reasoning via a Latent-Space Unified Model
460. NeurIPS 2026LeAct: Learning to Reason from Expert Actions
461. NeurIPS 2026LEAD: Length-Efficient Adaptive and Dynamic Reasoning for Large Language Models
462. NeurIPS 2026Learning to Complete Extremal Mathematical Structures
463. NeurIPS 2026Learning When to Think: Dual-Reference Offline Optimization for Adaptive VLM Reasoning
464. NeurIPS 2026LLM Agents Already Know When to Call Tools - Even Without Reasoning
465. NeurIPS 2026LLMs Show No Signs Of Individuated Metacognition
466. NeurIPS 2026Look Before You Reason: Implicit Visual Thinking for Efficient Multimodal Reasoning
467. NeurIPS 2026Low-Rank Hierarchical Merging for Efficient Long-to-Short Reasoning
468. NeurIPS 2026MammoGPS: A Benchmark for Visual Grounding, Perception, and Spatial Reasoning in Mammography
469. NeurIPS 2026Masked Diffusion Language Agents for Tool-Integrated Chemical Reasoning
470. NeurIPS 2026MCPHallu: Benchmarking Reasoning, Execution, and Memory Hallucinations in MCP Agents
471. NeurIPS 2026Measuring Weak-to-Strong Legibility of Reasoning Models
472. NeurIPS 2026MedZERO: Self-Evolving Agents for Open-Ended Medical Reasoning Through Controlled Knowledge Accumulation
473. NeurIPS 2026MemPilot: Learning Transferable Latent Memory Mechanisms for LLM Reasoning
474. NeurIPS 2026Mental Imagery That Matters: Curing Latent Collapse in Multimodal Reasoning
475. NeurIPS 2026MindLoom: Composing Thought Modes for Frontier-Level Reasoning Data Synthesis
476. NeurIPS 2026MinPath: Learning Efficient LLM Reasoning via Minimal Dependency Paths
477. NeurIPS 2026MIRA: Reinforcing Multimodal Reasoning via Deceptive Contextual Augmentation
478. NeurIPS 2026MIRAGE: Mobile Agents with Implicit Reasoning and Generative World Models
479. NeurIPS 2026Mitigating Factual Hallucination in Large Reasoning Models via Mixed-Mode Advantage Regularization
480. NeurIPS 2026Mitigating Overthinking in Large Reasoning Language Models via Reasoning Path Deviation Monitoring
481. NeurIPS 2026MMFineReason: Closing the Multimodal Reasoning Gap via Open Data-Centric Methods
482. NeurIPS 2026MotionHalluc: Diagnosing Kinematic Hallucinations in Fine-Grained Motion Reasoning
483. NeurIPS 2026MUX: Continuous Reasoning via Multiplexed Tokens
484. NeurIPS 2026Natural Synthesis: Outperforming Reactive Synthesis Tools with Large Reasoning Models
485. NeurIPS 2026On the Token Value Inequality in Efficient Reasoning
486. NeurIPS 2026OS-Pruner: Pruning Chains-of-Thought of Reasoning Models via Optimal Stopping
487. NeurIPS 2026Outcome-Based RL Provably Leads Transformers to Reason, but Only With the Right Data
488. NeurIPS 2026PAC Reasoning: Controlling the Performance Loss for Efficient Reasoning
489. NeurIPS 2026Pause and Reflect: Conformal Aggregation for Chain-of-Thought Reasoning
490. NeurIPS 2026Perceive, Interact, Reason: Building Tool-Augmented Visual Agents for Spatial Reasoning
491. NeurIPS 2026PhysGraphNet: Physical-State Scene Graphs via Latent Graph Reasoning and Counterfactual Supervision
492. NeurIPS 2026PhysVista: Benchmarking Physical Intelligence in VLMs via a Perception-Reasoning-Assessment Loop
493. NeurIPS 2026Plausibility Is Not Prediction: Contrastive Evidence for LLM-Based Cellular Perturbation Reasoning
494. NeurIPS 2026Position: Robust Reasoning Requires Internal Time, Not Scale Alone
495. NeurIPS 2026Precise but Uncoupled: Reviewer Precision Does Not Guarantee Critique Uptake in Multi-Agent Math Reasoning
496. NeurIPS 2026Preserving Exploration for LLM Reasoning via Mean Order-Statistic Alignment
497. NeurIPS 2026Probabilistic Recursive Reasoning
498. NeurIPS 2026Probing the Trajectories of Reasoning Traces in Large Language Models
499. NeurIPS 2026Problem-Dependent Dynamic Regret over a Predictor Class with One-Gradient Feedback
500. NeurIPS 2026Programmatic Reasoning with Structural Schema: A Unified Framework for Multi-Table Inference

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。