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 1,799 篇，第 12 / 36 页

  1. ICLR 2026Not Search, But Scan: Benchmarking MLLMs on Scan-Oriented Academic Paper Reasoning
  2. ICLR 2026Nudging the Boundaries of LLM Reasoning
  3. ICLR 2026OCR-Reasoning Benchmark: Unveiling the True Capabilities of MLLMs in Complex Text-Rich Image Reasoning
  4. ICLR 2026ODI-Bench: Can MLLMs Understand Immersive Omnidirectional Environments?
  5. ICLR 2026Off-Trajectory Reasoning: Can LLMs Collaborate on Reasoning Trajectory?
  6. ICLR 2026Omni-Weather: Unified Multimodal Foundation Model for Weather Generation and Understanding
  7. ICLR 2026OmniEVA: Embodied Versatile Planner via Task-Adaptive 3D-Grounded and Embodiment-aware Reasoning
  8. ICLR 2026OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Models
  9. ICLR 2026OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs
  10. ICLR 2026On Code-Induced Reasoning in LLMs
  11. ICLR 2026On the Design of KL-Regularized Policy Gradient Algorithms for LLM Reasoning
  12. ICLR 2026On the Direction of RLVR Updates for LLM Reasoning: Identification and Exploitation
  13. ICLR 2026On the Eligibility of LLMs for Counterfactual Reasoning: A Decompositional Study
  14. ICLR 2026On The Fragility of Benchmark Contamination Detection in Reasoning Models
  15. ICLR 2026On the Reasoning Abilities of Masked Diffusion Language Models
  16. ICLR 2026On the Thinking-Language Modeling Gap in Large Language Models
  17. ICLR 2026OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoning
  18. ICLR 2026OpenEstimate: Evaluating LLMs on Probabilistic Estimation with Real-World Data
  19. ICLR 2026OpenThoughts: Data Recipes for Reasoning Models
  20. ICLR 2026Optimal Aggregation of LLM and PRM Signals for Efficient Test-Time Scaling
  21. ICLR 2026Optimal Sparsity of Mixture-of-Experts Language Models for Reasoning Tasks
  22. ICLR 2026OptimalThinkingBench: Evaluating Over and Underthinking in LLMs
  23. ICLR 2026Output Supervision Can Obfuscate the Chain of Thought
  24. ICLR 2026Overthinking Reduction with Decoupled Rewards and Curriculum Data Scheduling
  25. ICLR 2026OWL : Geometry-Aware Spatial Reasoning for Audio Large Language Models
  26. ICLR 2026Parallel-R1: Towards Parallel Thinking via Reinforcement Learning
  27. ICLR 2026ParoQuant: Pairwise Rotation Quantization for Efficient Reasoning LLM Inference
  28. ICLR 2026Patching Gaps In LLM Reasoning With Interventional Training
  29. ICLR 2026PathChat-SegR1: Reasoning Segmentation in Pathology via SO-GRPO
  30. ICLR 2026PEAR: Phase Entropy Aware Reward for Efficient Reasoning
  31. ICLR 2026Perception-Aware Policy Optimization for Multimodal Reasoning
  32. ICLR 2026Perception-R1: Advancing Multimodal Reasoning Capabilities of MLLMs via Visual Perception Reward
  33. ICLR 2026PERK: Long-Context Reasoning as Parameter-Efficient Test-Time Learning
  34. ICLR 2026Persona Features Control Emergent Misalignment
  35. ICLR 2026Personalized Reasoning: Just-in-time Personalization and Why LLMs Fail at It
  36. ICLR 2026PILOT-Bench: Probabilistic Interaction for LLM Operations in Tool-driven Scenarios
  37. ICLR 2026Plan and Budget: Effective and Efficient Test-Time Scaling on Reasoning Large Language Models
  38. ICLR 2026PM-KVQ: Progressive Mixed-precision KV Cache Quantization for Long-CoT LLMs
  39. ICLR 2026Polychromic Objectives for Reinforcement Learning
  40. ICLR 2026Predicting LLM Output Length via Entropy-Guided Representations
  41. ICLR 2026Predicting LLM Reasoning Performance with Small Proxy Model
  42. ICLR 2026PreferThinker: Reasoning-based Personalized Image Preference Assessment
  43. ICLR 2026Premise Selection for a Lean Hammer
  44. ICLR 2026Pretraining with hierarchical memories: separating long-tail and common knowledge
  45. ICLR 2026Prior-free Tabular Test-time Adaptation
  46. ICLR 2026PRISM-Physics: Causal DAG-Based Process Evaluation for Physics Reasoning
  47. ICLR 2026PrismAudio: Decomposed Chain-of-Thought and Multi-dimensional Rewards for Video-to-Audio Generation
  48. ICLR 2026Privacy-Protected Causal Survival Analysis Under Distribution Shift
  49. ICLR 2026Process-Verified Reinforcement Learning for Theorem Proving via Lean
  50. ICLR 2026Programming by Backprop: Learning Behaviour from Symbolic Descriptions

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