MCTS论文趋势 2023–2026

主要结论

• MCTS论文从 2023 年的 115 篇增长到 2026 年的 278 篇，三年间扩大到约 2.4 倍。
• NeurIPS 是发MCTS论文最多的会议，2023–2026 年共 302 篇，占四个会议总量的 39%。
• 交叉方向里涨得最快的是AI 智能体：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 26 篇。
• 在「博弈与规划」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的经典规划是 3.3 倍，MCTS是 2.4 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 Regularization-Based Approach to Public Belief State Search for Adversarial Games
• ICML 2026$\alpha$-PFN: Fast Entropy Search via In-Context Learning
• ICLR 2026Better Learning-Augmented Spanning Tree Algorithms via Metric Forest Completion
• CVPR 2026DRS-GUI Dynamic Region Search for Training-Free GUI Grounding
• NeurIPS 2025A Single-Swap Local Search Algorithm for k-Means of Lines
• ICML 2025Accelerating Large Language Model Reasoning via Speculative Search
• ICLR 2025Analysis of Linear Mode Connectivity via Permutation-Based Weight Matching: With Insights into Other Permutation Search Methods
• CVPR 2025DyFo: A Training-Free Dynamic Focus Visual Search for Enhancing LMMs in Fine-Grained Visual Understanding
• NeurIPS 2024Speculative Monte-Carlo Tree Search
• ICML 2024Accelerated Speculative Sampling Based on Tree Monte Carlo
• ICLR 2024Chain of Log-Concave Markov Chains
• CVPR 2024Hierarchical Correlation Clustering and Tree Preserving Embedding
• NeurIPS 2023Accelerating Monte Carlo Tree Search with Probability Tree State Abstraction
• ICML 2023A new near-linear time algorithm for k-nearest neighbor search using a compressed cover tree
• ICLR 2023A GNN-Guided Predict-and-Search Framework for Mixed-Integer Linear Programming
• CVPR 2023Bridging Search Region Interaction With Template for RGB-T Tracking
• NeurIPS 2026A Revisit of Hamiltonian Monte Carlo Efficiency on Bayesian Neural Networks
• ICML 2026A Bi-metric Framework for Efficient Nearest Neighbor Search
• ICLR 2026Dr.LLM: Dynamic Layer Routing in LLMs
• CVPR 2026UAST Unified Active Search and Tracking for Arbitrary Targets with

全部MCTS论文

共 251 篇，第 2 / 6 页

51. ICML 2026Latent Guided Sampling for Combinatorial Optimization
52. ICML 2026LEAP: Zone-Aware MCTS for LLM Self-Speculative Decoding
53. ICML 2026Learning $U$-Statistics with Active Inference
54. ICML 2026Learning to Search and Searching to Learn for Generalization in Planning
55. ICML 2026Lookahead Sample Reward Guidance for Test-Time Scaling of Diffusion Models
56. ICML 2026Markov Chain Monte Carlo without Evaluating the Target: an Auxiliary Variable Approach
57. ICML 2026MARS: Modular Agent with Reflective Search for Automated AI Research
58. ICML 2026MAST: Motif-Augmented Diffusion with Search Tree for Spectroscopic Molecular Structure Elucidation
59. ICML 2026Matrix-Free GPU Semidefinite Programming for Quantum Ordered Search at the k=6 Frontier
60. ICML 2026MFH-NAS:A Hybrid Neural Architecture Search Framework for Multimodal Fusion Object Detection
61. ICML 2026MINT: Minimal Information Neuro-Symbolic Tree for Objective-Driven Knowledge-Gap Reasoning and Active Elicitation
62. ICML 2026Mitigating Premature Exploitation in Particle-based Monte Carlo for Inference-Time Scaling
63. ICML 2026Model-Free Robust Average-Reward Reinforcement Learning with Sample Complexity Analysis
64. ICML 2026Neural Attention Search Linear: Towards Adaptive Token-Level Hybrid Attention Models
65. ICML 2026Neural Low-Discrepancy Sequences
66. ICML 2026Neural Quantum States in Mixed Precision
67. ICML 2026Noisy Pairwise-Comparison Random Search for Smooth Nonconvex Optimization
68. ICML 2026NonZero: Interaction-Guided Exploration for Multi-Agent Monte Carlo Tree Search
69. ICML 2026OC-space: a Unifying Perspective on Verification of Tree Ensembles
70. ICML 2026On the Power of (Approximate) Reward Models for Inference-Time Scaling: Sequential Monte Carlo and Beyond
71. ICML 2026OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search
72. ICML 2026Operator Splitting with Hamilton-Jacobi-based Proximals
73. ICML 2026Optimal Quantum Speedups for Repeatedly Nested Expectation Estimation
74. ICML 2026Path-dependent Discrete Amortized Inference
75. ICML 2026PINE: Pruning Boosted Tree Ensembles with Conformal In-Distribution Prediction Equivalence
76. ICML 2026Practical and Scalable Hamiltonian Monte Carlo Without the Metropolis Test
77. ICML 2026Probabilistic Bisection Algorithm Provably Achieves Exponential Convergence
78. ICML 2026Proximal-IMH: Proximal Posterior Proposals for Independent Metropolis–Hastings with Approximate Operators
79. ICML 2026Recursive Monte-Carlo Tree Search
80. ICML 2026RefChess: Training-Free Contextual Search for Zero-Shot Referring Image Segmentation
81. ICML 2026Reflective Hamiltonian Monte Carlo: Mixing Analysis and Application to Sampling on Stiefel Manifold
82. ICML 2026Reinforced Sequential Monte Carlo for Amortised Sampling
83. ICML 2026RL4RLA: Teaching ML to Discover Randomized Linear Algebra Algorithms Through Curriculum Design and Graph-Based Search
84. ICML 2026SCHUR-A*: Layer-wise Optimal Expert Pruning for MoEs via Schur-Complement Guided A* Search
85. ICML 2026Score-Repellent Monte Carlo: Toward Efficient Non-Markovian Sampler with Constant Memory in General State Spaces
86. ICML 2026Search or Accelerate: Confidence-Switched Position Beam Search for Diffusion Language Models
87. ICML 2026Search Space Synthesis for Parametric Functions
88. ICML 2026Self-Prophetic Decoding to Unlock Visual Search in LVLMs
89. ICML 2026Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures
90. ICML 2026Sonar-TS: Search-Then-Verify Natural Language Querying for Time Series Databases
91. ICML 2026Stationary MMD Points
92. ICML 2026Structure-Induced Information for Rerooting Levin Tree Search
93. ICML 2026Structured Progressive Knowledge Activation for LLM-Driven Neural Architecture Search
94. ICML 2026SURGE: Approximation and Training Free Particle Filter for Diffusion Surrogate
95. ICML 2026SVL: Goal-Conditioned Reinforcement Learning as Survival Learning
96. ICML 2026Test-Time Graph Search for Goal-Conditioned Reinforcement Learning
97. ICML 2026TFTF: Training-Free Targeted Flow for Conditional Sampling
98. ICML 2026The (Marginal) Value of a Search Ad: An Online Causal Framework for Repeated Second-price Auctions
99. ICML 2026The Surprising Difficulty of Search in Model-Based Reinforcement Learning
100. ICML 2026The Trojan Knowledge: Bypassing Commercial LLM Guardrails via Harmless Prompt Weaving and Adaptive Tree Search

常见问题

2023–2026 年四大会议一共有多少篇MCTS论文？

共 773 篇，按年份分别是 2023 年 115 篇、2024 年 144 篇、2025 年 236 篇、2026 年 278 篇。

NeurIPS 2026 有多少篇MCTS论文？

NeurIPS 2026 有 99 篇MCTS论文，上一届是 100 篇。

ICLR 2026 有多少篇MCTS论文？

ICLR 2026 有 60 篇MCTS论文，上一届是 49 篇。

ICML 2026 有多少篇MCTS论文？

ICML 2026 有 105 篇MCTS论文，上一届是 75 篇。

CVPR 2026 有多少篇MCTS论文？

CVPR 2026 有 14 篇MCTS论文，上一届是 12 篇。

哪个会议的MCTS论文最多？

NeurIPS，2023–2026 年共 302 篇。