Talks · Publications
Talks & Publications
Talks
-
Spectral Representation-based Reinforcement Learning
JMLRJournal of Machine Learning ResearchAccepted
-
Spectral Ghost in Representation Learning: from Component Analysis to Self-Supervised Learning
Selected Recent PublicationsFull list on Google Scholar →
Peer-reviewed publications
-
FlowRL: A Taxonomy and Modular Framework for Reinforcement Learning with Diffusion Policies
RLC 2026Reinforcement Learning Conference
-
Diffusion Controller: Framework, Algorithms and Parameterization
ICML 2026International Conference on Machine Learning
-
Representation Learning via Non-Contrastive Mutual Information
AISTATS 2026International Conference on Artificial Intelligence and StatisticsSpotlight
-
MLE-Smith: Scaling MLE Tasks with Automated Multi-agent Pipeline
ICLR 2026International Conference on Learning Representations
-
Spectral Bellman Method: Unifying Representation and Exploration in RL
ICLR 2026International Conference on Learning Representations
-
AmorLIP: Efficient Language-Image Pretraining via Amortization
NeurIPS 2025Advances in Neural Information Processing Systems
-
MLE-Dojo: Interactive Environments for Empowering LLM Agents in Machine Learning Engineering
NeurIPS 2025Advances in Neural Information Processing Systems
-
Exploration from a Primal-Dual Lens: Value-Incentivized Actor-Critic Methods for Sample-Efficient Online RL
NeurIPS 2025Advances in Neural Information Processing Systems
-
Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs
NeurIPS 2025Advances in Neural Information Processing Systems
-
Efficient Online Reinforcement Learning for Diffusion Policy
ICML 2025International Conference on Machine Learning
-
Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF
ICLR 2025International Conference on Learning Representations
-
Diffusion Spectral Representation for Reinforcement Learning
NeurIPS 2024Advances in Neural Information Processing Systems
-
BBox-Adapter: Lightweight Adapting for Black-Box Large Language Models
ICML 2024International Conference on Machine LearningSpotlight
-
Provable Representation with Efficient Planning for Partially Observable Reinforcement Learning
ICML 2024International Conference on Machine Learning
-
Learning Universal Policies via Text-Guided Video Generation
NeurIPS 2023Advances in Neural Information Processing SystemsSpotlight
-
AdaPlanner: Adaptive Planning from Feedback with Language Models
NeurIPS 2023Advances in Neural Information Processing Systems
-
Stochastic Gradient Succeeds for Bandits
ICML 2023International Conference on Machine Learning