推理模型论文趋势 2023–2026

链式推理、测试时计算与显式推理能力

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 3,512 篇被归入该主题的论文。

主要结论

每年论文数

各会议论文数

按年份和会议统计

推理模型论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
20232715710933470
2024121846343311
202550020315970932
20263616525822041,799

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

全部推理模型论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 913 篇，第 14 / 19 页

  1. ICML 2025LMAct: A Benchmark for In-Context Imitation Learning with Long Multimodal Demonstrations
  2. ICML 2025Lower Bounds for Chain-of-Thought Reasoning in Hard-Attention Transformers
  3. ICML 2025MA-LoT: Model-Collaboration Lean-based Long Chain-of-Thought Reasoning enhances Formal Theorem Proving
  4. ICML 2025Machine Learning meets Algebraic Combinatorics: A Suite of Datasets Capturing Research-level Conjecturing Ability in Pure Mathematics
  5. ICML 2025Machines and Mathematical Mutations: Using GNNs to Characterize Quiver Mutation Classes
  6. ICML 2025MapEval: A Map-Based Evaluation of Geo-Spatial Reasoning in Foundation Models
  7. ICML 2025MARGE: Improving Math Reasoning with Guided Exploration
  8. ICML 2025MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations
  9. ICML 2025MathConstruct: Challenging LLM Reasoning with Constructive Proofs
  10. ICML 2025Metastable Dynamics of Chain-of-Thought Reasoning: Provable Benefits of Search, RL and Distillation
  11. ICML 2025Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worse
  12. ICML 2025Mirror, Mirror of the Flow: How Does Regularization Shape Implicit Bias?
  13. ICML 2025MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency
  14. ICML 2025Neural Genetic Search in Discrete Spaces
  15. ICML 2025Noise Conditional Variational Score Distillation
  16. ICML 2025On the Power of Context-Enhanced Learning in LLMs
  17. ICML 2025One Example Shown, Many Concepts Known! Counterexample-Driven Conceptual Reasoning in Mathematical LLMs
  18. ICML 2025Optimizing Test-Time Compute via Meta Reinforcement Finetuning
  19. ICML 2025PDE-Controller: LLMs for Autoformalization and Reasoning of PDEs
  20. ICML 2025PENCIL: Long Thoughts with Short Memory
  21. ICML 2025PokéChamp: an Expert-level Minimax Language Agent
  22. ICML 2025Policy Guided Tree Search for Enhanced LLM Reasoning
  23. ICML 2025Position: A Theory of Deep Learning Must Include Compositional Sparsity
  24. ICML 2025Position: Formal Mathematical Reasoning—A New Frontier in AI
  25. ICML 2025Position: It Is Time We Test Neural Computation In Vitro
  26. ICML 2025Position: Language model developers should report train-test overlap
  27. ICML 2025Position: LLMs Need a Bayesian Meta-Reasoning Framework for More Robust and Generalizable Reasoning
  28. ICML 2025Position: Trustworthy AI Agents Require the Integration of Large Language Models and Formal Methods
  29. ICML 2025Position: We Need An Algorithmic Understanding of Generative AI
  30. ICML 2025Premise-Augmented Reasoning Chains Improve Error Identification in Math reasoning with LLMs
  31. ICML 2025Primal-Dual Neural Algorithmic Reasoning
  32. ICML 2025Product of Experts with LLMs: Boosting Performance on ARC Is a Matter of Perspective
  33. ICML 2025PTTA: Purifying Malicious Samples for Test-Time Model Adaptation
  34. ICML 2025Putnam-AXIOM: A Functional & Static Benchmark for Measuring Higher Level Mathematical Reasoning in LLMs
  35. ICML 2025R2-T2: Re-Routing in Test-Time for Multimodal Mixture-of-Experts
  36. ICML 2025Ranked Entropy Minimization for Continual Test-Time Adaptation
  37. ICML 2025RE-IMAGINE: Symbolic Benchmark Synthesis for Reasoning Evaluation
  38. ICML 2025Reasoning Limitations of Multimodal Large Language Models. A case study of Bongard Problems
  39. ICML 2025Reasoning Through Execution: Unifying Process and Outcome Rewards for Code Generation
  40. ICML 2025Reasoning-as-Logic-Units: Scaling Test-Time Reasoning in Large Language Models Through Logic Unit Alignment
  41. ICML 2025ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding
  42. ICML 2025Regress, Don't Guess: A Regression-like Loss on Number Tokens for Language Models
  43. ICML 2025RepoAudit: An Autonomous LLM-Agent for Repository-Level Code Auditing
  44. ICML 2025Rethinking Chain-of-Thought from the Perspective of Self-Training
  45. ICML 2025Rethinking External Slow-Thinking: From Snowball Errors to Probability of Correct Reasoning
  46. ICML 2025ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verification
  47. ICML 2025Revisiting Chain-of-Thought in Code Generation: Do Language Models Need to Learn Reasoning before Coding?
  48. ICML 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking
  49. ICML 2025RULEBREAKERS: Challenging LLMs at the Crossroads between Formal Logic and Human-like Reasoning
  50. ICML 2025Safety Reasoning with Guidelines

在 Claude、Cursor 里直接问这个趋势

接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。

可以直接复制这句话问 AI

用 paperlist 查推理模型在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？

AI 会调用的工具：get_topic_stats {"topic": "reasoning-models"}

两分钟接入 Paperlist MCP（Claude Code、Cursor、Codex、ChatGPT）

常见问题

2023–2026 年四大会议一共有多少篇推理模型论文？
共 3,512 篇，按年份分别是 2023 年 470 篇、2024 年 311 篇、2025 年 932 篇、2026 年 1,799 篇。
NeurIPS 2026 有多少篇推理模型论文？
NeurIPS 2026 有 361 篇推理模型论文，上一届是 500 篇。
ICLR 2026 有多少篇推理模型论文？
ICLR 2026 有 652 篇推理模型论文，上一届是 203 篇。
ICML 2026 有多少篇推理模型论文？
ICML 2026 有 582 篇推理模型论文，上一届是 159 篇。
CVPR 2026 有多少篇推理模型论文？
CVPR 2026 有 204 篇推理模型论文，上一届是 70 篇。
哪个会议的推理模型论文最多？
NeurIPS，2023–2026 年共 1,253 篇。

「前沿系统」下的其他主题