人形控制论文趋势 2023–2026

主要结论

• 人形控制论文从 2023 年的 59 篇增长到 2026 年的 226 篇，三年间扩大到约 3.8 倍。
• NeurIPS 是发人形控制论文最多的会议，2023–2026 年共 188 篇，占四个会议总量的 40%。
• 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 1 篇增至 2026 年的 31 篇。
• 在「机器人与具身」的 5 个主题里，它的 2026 年论文数排第 2；同组增长最快的机器人操作是 3.1 倍，人形控制是 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 Control-Theoretic Approximation to Predictive Coding Dynamics
• ICML 2026A Control-Theoretic View of Mamba on Stability and Robustness
• ICLR 2026A Schrödinger Eigenfunction Method for Long-Horizon Stochastic Optimal Control
• CVPR 2026ActAvatar Temporally-Aware Precise Action Control for Talking Avatars
• NeurIPS 2025Improving Bilinear RNN with Closed-loop Control
• ICML 2025Learn to Vaccinate: Combining Structure Learning and Effective Vaccination for Epidemic and Outbreak Control
• ICLR 2025MamKO: Mamba-based Koopman operator for modeling and predictive control
• CVPR 2025ControlFace: Harnessing Facial Parametric Control for Face Rigging
• NeurIPS 2024Humanoid Locomotion as Next Token Prediction
• ICML 2024Adapting Pretrained ViTs with Convolution Injector for Visuo-Motor Control
• ICLR 2024Conformal Risk Control
• CVPR 2024Egocentric Whole-Body Motion Capture with FisheyeViT and Diffusion-Based Motion Refinement
• NeurIPS 2023AMDP: An Adaptive Detection Procedure for False Discovery Rate Control in High-Dimensional Mediation Analysis
• ICML 2023Generating Language Corrections for Teaching Physical Control Tasks
• ICLR 2023Efficient Attention via Control Variates
• CVPR 2023DC2: Dual-Camera Defocus Control by Learning To Refocus
• NeurIPS 2026Anytime-Valid Conformal Risk Control
• ICML 2026A Two-Layer Framework for Joint Online Configuration Selection and Admission Control
• ICLR 2026Causal-Steer: Disentangled Continuous Style Control without Parallel Corpora
• CVPR 2026ChArtist Generating Pictorial Charts with Unified Spatial and Subject Control

全部人形控制论文

共 475 篇，第 3 / 10 页

101. ICML 2026Error Propagation in Dynamic Programming: From Stochastic Control to American Option Pricing
102. ICML 2026Fast Mixing Steady-State Control in Markov Decision Processes
103. ICML 2026Generalized Boundary FDR Control under Arbitrary Dependence: An Approach on Closure Principle
104. ICML 2026Geometric Control of Out-of-Distribution Shift in Safe Offline RL
105. ICML 2026Knowledge Diversion for Efficient Morphology Control and Policy Transfer
106. ICML 2026Learning to Perceive the World Through Control: Empowerment-Based Representation Learning
107. ICML 2026Learning to Reconfigure: Configuration-Control Co-optimization of Reconfigurable Robots for Heterogeneous Locomotion
108. ICML 2026Learning Transferable Interaction Primitives from Game Videos for Humanoid Locomotion
109. ICML 2026Learning Treatment Allocations with Risk Control Under Partial Identifiability
110. ICML 2026LEC: Linear Expectation Constraints for Selection-Conditioned Risk Control in Selective Prediction and Routing Systems
111. ICML 2026Letting Trajectories Spread: Quality-Preserving Control for Diverse Flow Matching
112. ICML 2026Manifold-Optimal Guidance: A Unified Riemannian Control View of Diffusion Guidance
113. ICML 2026Model-Based Diffusion Sampling for Predictive Control in Offline Decision Making
114. ICML 2026Modular Pretraining Enables Access Control
115. ICML 2026No More, No Less: Least-Privilege Language Models
116. ICML 2026OGPO: Sample Efficient Full-Finetuning of Generative Control Policies
117. ICML 2026Optimal Transport–Guided Stochastic Control for Graph Combinatorial Optimization
118. ICML 2026Posterior Sampling Reinforcement Learning with Gaussian Processes for Continuous Control: Sublinear Regret Bounds for Unbounded State Spaces
119. ICML 2026Practical Mechanism for Fault-Tolerant Spiking Neural Networks via Simple Input Control Based on Learnable Fragmentation
120. ICML 2026ProMeCD: Unifying Long-Tailed and Noisy Label Learning via White-Box Control
121. ICML 2026RACER: Risk-Aware Calibrated Efficient Routing for Large Language Models
122. ICML 2026RBCBF: Decoding Time Safety Alignment via Risk Guided Rollback and Barrier Control
123. ICML 2026Reward Shaping Control Variates for Off-Policy Evaluation Under Sparse Rewards
124. ICML 2026Scalable and General Whole-Body Control for Cross-Humanoid Locomotion
125. ICML 2026ScaleMoE: Mixture-of-Experts for Scalable Continuous Control in Actor-Critic Reinforcement Learning
126. ICML 2026SCORE: A Unified Framework for Overshoot Refund in Online FDR Control
127. ICML 2026Solving Inverse Problems with Flow-based Models via Model Predictive Control
128. ICML 2026Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control
129. ICML 2026STAR-KV: Low-Rank KV Cache Compression via Soft Thresholding for Adaptive Rank Control
130. ICML 2026TetraJet-v2: Accurate NVFP4 Training for Large Language Models with Oscillation Suppression and Outlier Control
131. ICML 2026Wasserstein Geometry-Aware Adaptive Control via Meta-Learning
132. ICML 2026Well-Posed KL-Regularized Control via Wasserstein and Kalman–Wasserstein KL Divergences
133. NeurIPS 2026A Control-Theoretic Approximation to Predictive Coding Dynamics
134. NeurIPS 2026A latent control model for realistic rodent motion
135. NeurIPS 2026A Mean-Field Framework for Inference-Time Distributional Control of Diffusion Models
136. NeurIPS 2026A Surrogate Perspective on Convergence of Fixed-Target DQN
137. NeurIPS 2026ABC-Align: Prediction-Powered Alignment with Adaptive Bias Control
138. NeurIPS 2026Act, Validate, Adapt: Closing the Causal Discovery-Control Loop
139. NeurIPS 2026AdKnob: Ad Intensity Control and Labeling for LLM-Native Advertising
140. NeurIPS 2026AdvJudge-Zero: Binary Decision Flips in LLM-as-a-Judge via Adversarial Control Tokens
141. NeurIPS 2026AI Control for Sandbagging on Fuzzy Tasks
142. NeurIPS 2026Anytime-Valid Conformal Risk Control
143. NeurIPS 2026Bellman Residual Minimization for Control: Geometry, Stationarity, and Convergence
144. NeurIPS 2026Calibration Is Not Control: Intervention Advantage for LLM-Agent Oversight
145. NeurIPS 2026Can Entry-Wise Clipping Give Spectral Control of Stochastic Gradients?
146. NeurIPS 2026Causal Discovery with False Positive Error Control
147. NeurIPS 2026CCTimeBoost: Learning Time-Varying Relative Risk with Case-Control Boosted Trees
148. NeurIPS 2026Conformal Selective Acting: Anytime-Valid Risk Control for RLVR-Trained LLMs
149. NeurIPS 2026CorridorLight: Cooperation as Task Negotiation with Causal Gating for Traffic Signal Control
150. NeurIPS 2026DARPAN: Controllability-Aware Residual Filtering for Pretrained Physical Control

常见问题

2023–2026 年四大会议一共有多少篇人形控制论文？

共 475 篇，按年份分别是 2023 年 59 篇、2024 年 98 篇、2025 年 92 篇、2026 年 226 篇。

NeurIPS 2026 有多少篇人形控制论文？

NeurIPS 2026 有 94 篇人形控制论文，上一届是 38 篇。

ICLR 2026 有多少篇人形控制论文？

ICLR 2026 有 33 篇人形控制论文，上一届是 16 篇。

ICML 2026 有多少篇人形控制论文？

ICML 2026 有 54 篇人形控制论文，上一届是 23 篇。

CVPR 2026 有多少篇人形控制论文？

CVPR 2026 有 45 篇人形控制论文，上一届是 15 篇。

哪个会议的人形控制论文最多？

NeurIPS，2023–2026 年共 188 篇。