推理模型论文趋势 2023–2026

主要结论

• 推理模型论文从 2023 年的 470 篇增长到 2026 年的 1,799 篇，三年间扩大到约 3.8 倍。
• NeurIPS 是发推理模型论文最多的会议，2023–2026 年共 1,253 篇，占四个会议总量的 36%。
• 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 9 篇增至 2026 年的 437 篇。
• 在「前沿系统」的 4 个主题里，它的 2026 年论文数排第 3；同组增长最快的AI 智能体是 12 倍，推理模型是 3.8 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 Mechanistic Analysis of Looped Reasoning Language Models
• ICML 2026A Formal Comparison Between Chain of Thought and Latent Thought
• ICLR 2026A Simple "Motivation" Can Enhance Reinforcement Finetuning of Large Reasoning Models
• CVPR 2026Asking like Socrates Socrates helps VLMs understand remote sensing images
• NeurIPS 2025A Frustratingly Simple Yet Highly Effective Attack Baseline: Over 90% Success Rate Against the Strong Black-box Models of GPT-4.5/4o/o1
• ICML 2025A Reasoning-Based Approach to Cryptic Crossword Clue Solving
• ICLR 2025A Formal Framework for Understanding Length Generalization in Transformers
• CVPR 2025Argus: Vision-Centric Reasoning with Grounded Chain-of-Thought
• NeurIPS 2024Are Large Language Models Good Statisticians?
• ICML 2024Algorithm of Thoughts: Enhancing Exploration of Ideas in Large Language Models
• ICLR 2024Amortizing intractable inference in large language models
• CVPR 2024Generate Subgoal Images before Act: Unlocking the Chain-of-Thought Reasoning in Diffusion Model for Robot Manipulation with Multimodal Prompts
• NeurIPS 2023A benchmark of categorical encoders for binary classification
• ICML 2023Can We Scale Transformers to Predict Parameters of Diverse ImageNet Models?
• ICLR 2023Automatic Chain of Thought Prompting in Large Language Models
• CVPR 2023ActMAD: Activation Matching To Align Distributions for Test-Time-Training
• NeurIPS 2026A Solvable Model of Chain-of-Thought in In-Context Learning
• ICML 2026A Narrowing Geometry in Contaminated Reasoning
• ICLR 2026A State-Transition Framework for Efficient LLM Reasoning
• CVPR 2026Beyond Weak Supervision MLLMs-Guided Graded Knowledge Distillation for Unsupervised Camouflaged

全部推理模型论文

共 582 篇，第 3 / 12 页

101. ICML 2026Coupled Variational Reinforcement Learning for Language Model General Reasoning
102. ICML 2026CPMöbius: Iterative Coach–Player Reasoning for Data-Free Reinforcement Learning
103. ICML 2026Critique-Guided Distillation for Robust Reasoning via Refinement
104. ICML 2026CyberJurors: A Multi-Agent Simulation Task for E-Commerce Disputes Verdict
105. ICML 2026D-CORE: Incentivizing Task Decomposition in Large Reasoning Models for Complex Tool Use
106. ICML 2026D-FUSEr: Diverse Failure, Unified Success via Error-Distribution Shaping in LLM Reasoning
107. ICML 2026D²Evo: Dual Difficulty-Aware Self-Evolution for Data-Efficient Reinforcement Learning
108. ICML 2026Data Provenance Auditing of Fine-Tuned Large Language Models with a Text-Preserving Technique
109. ICML 2026DC-W2S: Dual-Consensus Weak-to-Strong Training for Reliable Process Reward Modeling in Biological Reasoning
110. ICML 2026DecepChain: Inducing Deceptive Reasoning in Large Language Models
111. ICML 2026Decompose and Recompose: Reasoning New Skills from Existing Abilities for Cross-Task Robotic Manipulation
112. ICML 2026Decomposed On-Policy Distillation for Vision-Language Reasoning: Steering Gradients for Visual Grounding
113. ICML 2026Deep Scientific Reasoning under Physical Constraints: Structure-Aware Spectrum Prediction
114. ICML 2026Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoning
115. ICML 2026DenseSteer: Steering Small Language Models towards Dense Math Reasoning
116. ICML 2026Depth-Breadth Synergy in RLVR: Unlocking LLM Reasoning Gains with Adaptive Exploration
117. ICML 2026Differential Smoothing Mitigates Sharpening and Improves LLM Reasoning
118. ICML 2026Directly Optimizing Natural Language Explanations for Behavioral Faithfulness: Simulatability and Recoverability
119. ICML 2026Discrete Adjoint Schrödinger Bridge Sampler
120. ICML 2026Discrete Diffusion with Physical Mass Constraints for \emph{De Novo} Peptide Sequencing
121. ICML 2026Discriminative Visual Process Rewards for Scaling Thinking at Test-Time with Images
122. ICML 2026DisPPO: Quantile-Based Distributional Reinforcement Learning for Large Language Models
123. ICML 2026Distilling Neuro-Symbolic Programs into 3D Multi-modal LLMs
124. ICML 2026Distilling Task-Level Coordination Policies for Generalizable Multi-Agent Cooperation
125. ICML 2026Diversity Matters: Revisiting Test-Time Compute in Vision-Language Models
126. ICML 2026Diversity Over Frequency: Rethinking Tool Use in Visual Chain-of-Thought Agents
127. ICML 2026Do LLMs Signal When They’re Right? Evidence from Neuron Agreement
128. ICML 2026Do Sparse Autoencoders Identify Reasoning Features in Language Models?
129. ICML 2026Do Vision and Text Cues Exhibit Evidential Coupling? UFO: A Benchmark for Compositional Multimodal Reasoning in Unified Models
130. ICML 2026DocHop: Benchmarking Out-of-domain Multi-hop Reasoning in Information-Dense Documents
131. ICML 2026DocVAL: Validated Chain-of-Thought Distillation for Grounded Document VQA
132. ICML 2026Does Math Reasoning Improve General LLM Capabilities? Understanding Transferability of LLM Reasoning
133. ICML 2026Does Reasoning Improve Seeing? Understanding When Vision-Language Models Benefit from Thinking
134. ICML 2026Does Your Reasoning Model Implicitly Know When to Stop Thinking?
135. ICML 2026Don't Force the Fit: Bounded Log-Likelihood Loss for Enhanced Reasoning in Large Language Models
136. ICML 2026Don't Ignore the Tail: Decoupled Distillation Produces Top Maths Students on an Academic Budget
137. ICML 2026DREAM-R: Multimodal Speculative Reasoning with RL-Based Refined Drafting, Precise Verification, and Fully Parallel Execution
138. ICML 2026DRIVE: Best Data Scheduling Practices for Reinforcement Learning with Verifiable Reward in Competitive Code Generation
139. ICML 2026DTS: Enhancing Large Reasoning Models via Decoding Tree Sketching
140. ICML 2026Dual-Latent Memory Routing for Vision-Language Reasoning
141. ICML 2026DyCon: Dynamic Reasoning Control via Evolving Difficulty Modeling
142. ICML 2026Dynamic Thinking-Token Selection for Efficient Reasoning in Large Reasoning Models
143. ICML 2026Dynamics Reveals Structure: Challenging the Linear Propagation Assumption
144. ICML 2026Dynamics Within Latent Chain-of-Thought: An Empirical Study of Causal Structure
145. ICML 2026DynVLA: Learning World Dynamics for Action Reasoning in Autonomous Driving
146. ICML 2026E-VAds: An E-commerce Short Videos Understanding Benchmark for MLLMs
147. ICML 2026EAGer: Entropy-Aware GEneRation for Adaptive Inference-Time Scaling
148. ICML 2026Early Decisions Matter: Proximity Bias and Initial Trajectory Shaping in Non-Autoregressive Diffusion Language Models
149. ICML 2026ECCO: Evidence-Driven Causal Reasoning for Compiler Optimization
150. ICML 2026Echoes within the Reasoning: Stealthy and Effective Watermarking via Chain of Thought

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。