通用机器学习 · reinforcement learning论文趋势 2023–2026

主要结论

• 通用机器学习 · reinforcement learning论文从 2023 年的 287 篇增长到 2026 年的 575 篇，三年间扩大到约 2.0 倍。
• ICLR 是发通用机器学习 · reinforcement learning论文最多的会议，2023–2026 年共 847 篇，占四个会议总量的 51%。
• 交叉方向里涨得最快的是扩散模型：同时属于两个主题的论文从 2023 年的 2 篇增至 2026 年的 34 篇。
• 在「通用机器学习」的 6 个主题里，它的 2026 年论文数排第 2；同组增长最快的通用机器学习 · evaluation是 4.3 倍，通用机器学习 · reinforcement learning是 2.0 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 2026The Sample Complexity of Multiple Change Point Identification under Bandit Feedback
• ICML 2026De-attribute to Forget for LLM Unlearning
• ICLR 2026AbstRaL: Augmenting LLMs' Reasoning by Reinforcing Abstract Thinking
• CVPR 2026Dynamic Important Example Mining for Reinforcement Finetuning
• NeurIPS 2025RULE: Reinforcement UnLEarning Achieves Forget-retain Pareto Optimality
• ICML 2025The Number of Trials Matters in Infinite-Horizon General-Utility Markov Decision Processes
• ICLR 2025A General Framework for Off-Policy Learning with Partially-Observed Reward
• CVPR 2025Automated Proof of Polynomial Inequalities via Reinforcement Learning
• NeurIPS 2024Learning General Parameterized Policies for Infinite Horizon Average Reward Constrained MDPs via Primal-Dual Policy Gradient Algorithm
• ICML 2024Provable Risk-Sensitive Distributional Reinforcement Learning with General Function Approximation
• ICLR 2024Bandits with Replenishable Knapsacks: the Best of both Worlds
• CVPR 2024Improving Unsupervised Hierarchical Representation with Reinforcement Learning
• NeurIPS 2023General Munchausen Reinforcement Learning with Tsallis Kullback-Leibler Divergence
• ICML 2023Non-stationary Reinforcement Learning under General Function Approximation
• ICLR 2023A General Framework for Sample-Efficient Function Approximation in Reinforcement Learning
• CVPR 2023Frustratingly Easy Regularization on Representation Can Boost Deep Reinforcement Learning
• NeurIPS 2026Learning When to Stop: Selective Imitation Learning Under Arbitrary Dynamics Shift
• ICML 2026Fox in the Henhouse: Supply-Chain Backdoor Attacks Against Reinforcement Learning
• ICLR 2026APPLE: Toward General Active Perception via Reinforcement Learning
• CVPR 2026Reinforcement-Guided Synthetic Data Generation for Privacy-Sensitive Identity Recognition

全部通用机器学习 · reinforcement learning论文

共 575 篇，第 7 / 12 页

301. ICLR 2026Spectral Bellman Method: Unifying Representation and Exploration in RL
302. ICLR 2026SPELL: Self-Play Reinforcement Learning for Evolving Long-Context Language Models
303. ICLR 2026SSVPO: Effective Step-Level Credit Assignment for RL Training of Language Models
304. ICLR 2026Stackelberg Coupling of Online Representation Learning and Reinforcement Learning
305. ICLR 2026Stackelberg Learning from Human Feedback: Preference Optimization as a Sequential Game
306. ICLR 2026STAIRS-Former: Spatio-Temporal Attention with Interleaved Recursive Structure TransFormer for Offline Mulit-task Multi-agent Reinforcement Learning
307. ICLR 2026Statistical Guarantees for Offline Domain Randomization
308. ICLR 2026Stop Unnecessary Reflection: Training LRMs for Efficient Reasoning with Adaptive Reflection and Length Coordinated Penalty
309. ICLR 2026Strict Subgoal Execution: Reliable Long-Horizon Planning in Hierarchical Reinforcement Learning
310. ICLR 2026Structured Reasoning for LLMs: A Unified Framework for Efficiency and Explainability
311. ICLR 2026SUSD: Structured Unsupervised Skill Discovery through State Factorization
312. ICLR 2026Tackling Heavy-Tailed Q-Value Bias in Offline-to-Online Reinforcement Learning with Laplace-Robust Modeling
313. ICLR 2026Task Tokens: A Flexible Approach to Adapting Behavior Foundation Models
314. ICLR 2026TD-JEPA: Latent-predictive Representations for Zero-Shot Reinforcement Learning
315. ICLR 2026Text2Grad: Reinforcement Learning from Natural Language Feedback
316. ICLR 2026The Choice of Divergence: A Neglected Key to Mitigating Diversity Collapse in Reinforcement Learning with Verifiable Reward
317. ICLR 2026The Rank and Gradient Lost in Non-stationarity: Sample Weight Decay for Mitigating Plasticity Loss in Reinforcement Learning
318. ICLR 2026The State of Reinforcement Finetuning for Transformer-based Generative Agents
319. ICLR 2026TimeSearch-R: Adaptive Temporal Search for Long-Form Video Understanding via Self-Verification Reinforcement Learning
320. ICLR 2026TIPS: Turn-level Information-Potential Reward Shaping for Search-Augmented LLMs
321. ICLR 2026Toward Conservative Planning from Preferences in Offline Reinforcement Learning
322. ICLR 2026Toward Efficient Exploration by Large Language Model Agents
323. ICLR 2026Towards Bridging the Gap between Large-Scale Pretraining and Efficient Finetuning for Humanoid Control
324. ICLR 2026Towards Efficient Constraint Handling in Neural Solvers for Routing Problems
325. ICLR 2026Towards High Data Efficiency in Reinforcement Learning with Verifiable Reward
326. ICLR 2026TRACED: Transition-aware Regret Approximation with Co-learnability for Environment Design
327. ICLR 2026Trajectory Generation with Conservative Value Guidance for Offline Reinforcement Learning
328. ICLR 2026Transitive RL: Value Learning via Divide and Conquer
329. ICLR 2026Triple-BERT: Do We Really Need MARL for Order Dispatch on Ride-Sharing Platforms?
330. ICLR 2026TROLL: Trust Regions Improve Reinforcement Learning for Large Language Models
331. ICLR 2026Trust-Region Adaptive Policy Optimization
332. ICLR 2026Uncovering Robot Vulnerabilities through Semantic Potential Fields
333. ICLR 2026Understanding and Improving Hyperbolic Deep Reinforcement Learning
334. ICLR 2026Universal Value-Function Uncertainties
335. ICLR 2026Unraveling the Complexity of Memory in RL Agents: an Approach for Classification and Evaluation
336. ICLR 2026Unsupervised Learning of Efficient Exploration: Pre-training Adaptive Policies via Self-Imposed Goals
337. ICLR 2026Use the Online Network If You Can: Towards Fast and Stable Reinforcement Learning
338. ICLR 2026Value Flows
339. ICLR 2026Value Gradient Flow: Behavior-Regularized RL without Regularization
340. ICLR 2026Variance-Dependent Regret Lower Bounds for Contextual Bandits
341. ICLR 2026VTool-R1: VLMs Learn to Think with Images via Reinforcement Learning on Multimodal Tool Use
342. ICLR 2026Wavelet Predictive Representations for Non-Stationary Reinforcement Learning
343. ICLR 2026WebArbiter: A Generative Reasoning Process Reward Model for Web Agents
344. ICLR 2026What Matters for Batch Online Reinforcement Learning in Robotics?
345. ICLR 2026Who Matters Matters: Agent-Specific Conservative Offline MARL
346. ICLR 2026WIMLE: Uncertainty‑Aware World Models with IMLE for Sample‑Efficient Continuous Control
347. ICLR 2026WorldGym: World Model as An Environment for Policy Evaluation
348. ICLR 2026XQC: Well-conditioned Optimization Accelerates Deep Reinforcement Learning
349. ICLR 2026Zero-Shot Adaptation of Behavioral Foundation Models to Unseen Dynamics
350. ICML 2026Action Manifold Smoothing: A Lipschitz Pathway Perspective on High-Dimensional Reinforcement Learning

常见问题

2023–2026 年四大会议一共有多少篇通用机器学习 · reinforcement learning论文？

共 1,676 篇，按年份分别是 2023 年 287 篇、2024 年 403 篇、2025 年 411 篇、2026 年 575 篇。

NeurIPS 2026 有多少篇通用机器学习 · reinforcement learning论文？

NeurIPS 2026 有 113 篇通用机器学习 · reinforcement learning论文，上一届是 108 篇。

ICLR 2026 有多少篇通用机器学习 · reinforcement learning论文？

ICLR 2026 有 335 篇通用机器学习 · reinforcement learning论文，上一届是 230 篇。

ICML 2026 有多少篇通用机器学习 · reinforcement learning论文？

ICML 2026 有 113 篇通用机器学习 · reinforcement learning论文，上一届是 71 篇。

CVPR 2026 有多少篇通用机器学习 · reinforcement learning论文？

CVPR 2026 有 14 篇通用机器学习 · reinforcement learning论文，上一届是 2 篇。

哪个会议的通用机器学习 · reinforcement learning论文最多？

ICLR，2023–2026 年共 847 篇。