强化学习论文趋势 2023–2026

主要结论

• 强化学习论文从 2023 年的 1,070 篇增长到 2026 年的 2,985 篇，三年间扩大到约 2.8 倍。
• NeurIPS 是发强化学习论文最多的会议，2023–2026 年共 2,494 篇，占四个会议总量的 36%。
• 交叉方向里涨得最快的是推理模型：同时属于两个主题的论文从 2023 年的 9 篇增至 2026 年的 437 篇。

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 2026\texttt{FEROM}: Frontier Endogenous Reveal-Order Marginal Policy Optimization for Masked Diffusion LMs
• ICML 2026$\text{DT}^\text{2}$: Decision-Targeted Digital Twins
• ICLR 2026$\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Textual Space
• CVPR 2026A Causal Marriage between VLM and IRM from Understanding to
• NeurIPS 2025$\texttt{G1}$: Teaching LLMs to Reason on Graphs with Reinforcement Learning
• ICML 2025A Computationally Efficient Algorithm for Infinite-Horizon Average-Reward Linear MDPs
• ICLR 2025$\phi$-Update: A Class of Policy Update Methods with Policy Convergence Guarantee
• CVPR 2025All-Optical Nonlinear Diffractive Deep Network for Ultrafast Image Denoising
• NeurIPS 2024A Critical Evaluation of AI Feedback for Aligning Large Language Models
• ICML 2024${\rm E}(3)$-Equivariant Actor-Critic Methods for Cooperative Multi-Agent Reinforcement Learning
• ICLR 2024A Black-box Approach for Non-stationary Multi-agent Reinforcement Learning
• CVPR 2024DMR: Decomposed Multi-Modality Representations for Frames and Events Fusion in Visual Reinforcement Learning
• NeurIPS 2023“Why Not Looking backward?” A Robust Two-Step Method to Automatically Terminate Bayesian Optimization
• ICML 2023A Connection between One-Step RL and Critic Regularization in Reinforcement Learning
• ICLR 2023$O(T^{-1})$ Convergence of Optimistic-Follow-the-Regularized-Leader in Two-Player Zero-Sum Markov Games
• CVPR 2023"Seeing" Electric Network Frequency From Events
• NeurIPS 2026$\varphi$TD: Distributional Reinforcement Learning using Characteristic Functions
• ICML 2026$\textit{S}$-SPPO: Semantic-Calibrated Self-Play Preference Optimization
• ICLR 2026$AutoDrive\text{-}P^3$: Unified Chain of Perception–Prediction–Planning Thought via Reinforcement Fine-Tuning
• CVPR 2026AdaptVision Efficient Vision-Language Models via Adaptive Visual Acquisition

全部强化学习论文

共 1,827 篇，第 11 / 37 页

501. ICLR 2026Nudging the Boundaries of LLM Reasoning
502. ICLR 2026Object-Centric World Models from Few-Shot Annotations for Sample-Efficient Reinforcement Learning
503. ICLR 2026Occupancy Reward Shaping: Improving Credit Assignment for Offline Goal-Conditioned Reinforcement Learning
504. ICLR 2026Octax: Accelerated CHIP-8 Arcade Environments for Reinforcement Learning in JAX
505. ICLR 2026Off-Policy Safe Reinforcement Learning with Cost-Constrained Optimistic Exploration
506. ICLR 2026Offline Preference-Based Value Optimization
507. ICLR 2026Offline Reinforcement Learning with Adaptive Feature Fusion
508. ICLR 2026On Entropy Control in LLM-RL Algorithms
509. ICLR 2026On Predictability of Reinforcement Learning Dynamics for Large Language Models
510. ICLR 2026On the Computational Limits of AI4S-RL : A Unified $\varepsilon$-$N$ Analysis
511. ICLR 2026On the Design of KL-Regularized Policy Gradient Algorithms for LLM Reasoning
512. ICLR 2026On the Direction of RLVR Updates for LLM Reasoning: Identification and Exploitation
513. ICLR 2026On The Fragility of Benchmark Contamination Detection in Reasoning Models
514. ICLR 2026On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification
515. ICLR 2026On the Tension Between Optimality and Adversarial Robustness in Policy Optimization
516. ICLR 2026On-Policy RL Meets Off-Policy Experts: Harmonizing Supervised Fine-Tuning and Reinforcement Learning via Dynamic Weighting
517. ICLR 2026One Life to Learn: Inferring Symbolic World Models for Stochastic Environments from Unguided Exploration
518. ICLR 2026One Model for All Tasks: Leveraging Efficient World Models in Multi-Task Planning
519. ICLR 2026One-Step Flow Q-Learning: Addressing the Diffusion Policy Bottleneck in Offline Reinforcement Learning
520. ICLR 2026OpenApps: Simulating Environment Variations to Measure UI Agent Reliability
521. ICLR 2026Operator Theory-Driven Autoformulation of MDPs for Control of Queueing Systems
522. ICLR 2026OPPO: Accelerating PPO-based RLHF via Pipeline Overlap
523. ICLR 2026Opponent Shaping in LLM Agents
524. ICLR 2026OPRIDE: Efficient Offline Preference-based Reinforcement Learning via In-Dataset Exploration
525. ICLR 2026Optimal Robust Subsidy Policies for Irrational Agent in Principal-Agent MDPs
526. ICLR 2026Optimal Sparsity of Mixture-of-Experts Language Models for Reasoning Tasks
527. ICLR 2026Optimistic Task Inference for Behavior Foundation Models
528. ICLR 2026OSIRIS: Bridging Analog Circuit Design and Machine Learning with Scalable Dataset Generation
529. ICLR 2026Overthinking Reduction with Decoupled Rewards and Curriculum Data Scheduling
530. ICLR 2026P-GenRM: Personalized Generative Reward Model with Test-time User-based Scaling
531. ICLR 2026PAMDP: Interact to Persona Alignment via a Partially Observable Markov Decision Process
532. ICLR 2026Parallel Multimodal Diffusion Language Models for Thinking-Aware Editing and Generation
533. ICLR 2026Parallel-R1: Towards Parallel Thinking via Reinforcement Learning
534. ICLR 2026Parameter-Efficient Reinforcement Learning using Prefix Optimization
535. ICLR 2026ParaS2S: Benchmarking and Aligning Spoken Language Models for Paralinguistic-aware Speech-to-Speech Interaction
536. ICLR 2026Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environments
537. ICLR 2026Patching Gaps In LLM Reasoning With Interventional Training
538. ICLR 2026Path Channels and Plan Extension Kernels: a Mechanistic Description of Planning in a Sokoban RNN
539. ICLR 2026PathChat-SegR1: Reasoning Segmentation in Pathology via SO-GRPO
540. ICLR 2026PCPO: Proportionate Credit Policy Optimization for Preference Alignment of Image Generation Models
541. ICLR 2026Peak-Return Greedy Slicing: Subtrajectory Selection for Transformer-based Offline RL
542. ICLR 2026Peng's Q($\lambda$) for Conservative Value Estimation in Offline Reinforcement Learning
543. ICLR 2026Perception-Aware Policy Optimization for Multimodal Reasoning
544. ICLR 2026Perception-R1: Advancing Multimodal Reasoning Capabilities of MLLMs via Visual Perception Reward
545. ICLR 2026Persona Features Control Emergent Misalignment
546. ICLR 2026Plan-R1: Safe and Feasible Trajectory Planning as Language Modeling
547. ICLR 2026Planning with an Embodied Learnable Memory
548. ICLR 2026Play to Generalize: Learning to Reason Through Game Play
549. ICLR 2026PLoP: Precise LoRA Placement for Efficient Finetuning of Large Models
550. ICLR 2026PoLi-RL: A Point-to-List Reinforcement Learning Framework for Conditional Semantic Textual Similarity

常见问题

2023–2026 年四大会议一共有多少篇强化学习论文？

共 6,885 篇，按年份分别是 2023 年 1,070 篇、2024 年 1,190 篇、2025 年 1,640 篇、2026 年 2,985 篇。

NeurIPS 2026 有多少篇强化学习论文？

NeurIPS 2026 有 760 篇强化学习论文，上一届是 784 篇。

ICLR 2026 有多少篇强化学习论文？

ICLR 2026 有 911 篇强化学习论文，上一届是 379 篇。

ICML 2026 有多少篇强化学习论文？

ICML 2026 有 1,021 篇强化学习论文，上一届是 405 篇。

CVPR 2026 有多少篇强化学习论文？

CVPR 2026 有 293 篇强化学习论文，上一届是 72 篇。

哪个会议的强化学习论文最多？

NeurIPS，2023–2026 年共 2,494 篇。