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 652 篇，第 12 / 14 页

  1. ICLR 2026THE PATH OF LEAST RESISTANCE: GUIDING LLM REASONING TRAJECTORIES WITH PREFIX CONSENSUS
  2. ICLR 2026The Quest for Efficient Reasoning: A Data-Centric Benchmark to CoT Distillation
  3. ICLR 2026The State of Reinforcement Finetuning for Transformer-based Generative Agents
  4. ICLR 2026THEMIS: Towards Holistic Evaluation of MLLMs for Scientific Paper Fraud Forensics
  5. ICLR 2026Theory of Space: Can Foundation Models Construct Spatial Beliefs through Active Exploration?
  6. ICLR 2026Think in Parallel, Answer as One: Logit Averaging for Open-Ended Reasoning
  7. ICLR 2026Think Then Embed: Generative Context Improves Multimodal Embedding
  8. ICLR 2026Think-While-Generating: On-the-Fly Reasoning for Personalized Long-Form Generation
  9. ICLR 2026Thinking as Society: Multi-Social-Agent Self-Distillation for Multimodal Misinformation Detection
  10. ICLR 2026Thinking on the Fly: Test-Time Reasoning Enhancement via Latent Thought Policy Optimization
  11. ICLR 2026Thinking-Free Policy Initialization Makes Distilled Reasoning Models More Effective and Efficient Reasoners
  12. ICLR 2026ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning
  13. ICLR 2026ThinkOmni: Lifting Textual Reasoning to Omni-modal Scenarios via Guidance Decoding
  14. ICLR 2026ThinKV: Thought-Adaptive KV Cache Compression for Efficient Reasoning Models
  15. ICLR 2026THOR: Tool-Integrated Hierarchical Optimization via RL for Mathematical Reasoning
  16. ICLR 2026Thought Branches: Interpreting LLM Reasoning Requires Resampling
  17. ICLR 2026Through the Lens of Contrast: Self-Improving Visual Reasoning in VLMs
  18. ICLR 2026TimeOmni-1: Incentivizing Complex Reasoning with Time Series in Large Language Models
  19. ICLR 2026TimeSearch-R: Adaptive Temporal Search for Long-Form Video Understanding via Self-Verification Reinforcement Learning
  20. ICLR 2026TimeSeriesExamAgent: Creating TimeSeries Reasoning Benchmarks at Scale
  21. ICLR 2026Tina: Tiny Reasoning Models via LoRA
  22. ICLR 2026TNT: Improving Chunkwise Training for Test-Time Memorization
  23. ICLR 2026Token-Efficient Long-Term Interest Sketching and Internalized Reasoning for LLM-based Recommendation
  24. ICLR 2026TokUR: Token-Level Uncertainty Estimation for Large Language Model Reasoning
  25. ICLR 2026Toward Effective Tool-Integrated Reasoning via Self-Evolved Preference Learning
  26. ICLR 2026Towards Faithful Reasoning in Remote Sensing: A Perceptually-Grounded GeoSpatial Chain-of-Thought for Vision-Language Models
  27. ICLR 2026Towards Knowledge‑and‑Data‑Driven Organic Reaction Prediction: RAG‑Enhanced and Reasoning‑Powered Hybrid System with LLMs
  28. ICLR 2026Towards Quantization-Aware Training for Ultra-Low-Bit Reasoning LLMs
  29. ICLR 2026Towards Safe Reasoning in Large Reasoning Models via Corrective Intervention
  30. ICLR 2026Traceable Evidence Enhanced Visual Grounded Reasoning: Evaluation and Method
  31. ICLR 2026Tracing the Traces: Latent Temporal Signals for Efficient and Accurate Reasoning
  32. ICLR 2026Trained on Tokens, Calibrated on Concepts: The Emergence of Semantic Calibration in LLMs
  33. ICLR 2026Training Large Language Models To Reason In Parallel With Global Forking Tokens
  34. ICLR 2026Training Large Reasoning Models Efficiently via Progressive Thought Encoding
  35. ICLR 2026Transductive Visual Programming: Evolving Tool Libraries from Experience for Spatial Reasoning
  36. ICLR 2026TraPO: A Semi-Supervised Reinforcement Learning Framework for Boosting LLM Reasoning
  37. ICLR 2026Tricks or Traps? A Deep Dive into RL for LLM Reasoning
  38. ICLR 2026TrimR: Verifier-based Training-Free Thinking Trimming for Efficient Test-Time Scaling
  39. ICLR 2026Trust-Region Adaptive Policy Optimization
  40. ICLR 2026Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanations
  41. ICLR 2026TSLM: Tree-Structured Language Modeling for Divergent Thinking
  42. ICLR 2026TUMIX: Multi-Agent Test-Time Scaling with Tool-Use Mixture
  43. ICLR 2026TumorChain: Interleaved Multimodal Chain-of-Thought Reasoning for Traceable Clinical Tumor Analysis
  44. ICLR 2026Type-Compliant Adaptation Cascades
  45. ICLR 2026UALM: Unified Audio Language Model for Understanding, Generation and Reasoning
  46. ICLR 2026UME-R1: Exploring Reasoning-Driven Generative Multimodal Embeddings
  47. ICLR 2026Understanding the Role of Training Data in Test-Time Scaling
  48. ICLR 2026Understanding VLMs Spatial Mental Modeling Capability from Limited Views
  49. ICLR 2026Understanding vs. Generation: Navigating Optimization Dilemma in Multimodal Models
  50. ICLR 2026Unfolding Spatial Cognition: Evaluating Multimodal Models on Visual Simulations

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