Dr Yali Du Academics Supervisors Head of the Distributed AI Research Group Senior Lecturer in Artificial Intelligence. Research subject areas Computer science Contact details yali.du@kcl.ac.uk
CODI: Compressing chain-of-thought into continuous space via self-distillation Ensemble Value Functions for Efficient Exploration in Multi-Agent Reinforcement Learning Goal-oriented Semantic Communication in Bandwidth-constrained MARL Integrating Large Language Models with Reinforcement Learning for Generalization in Strategic Card Games: Extended Abstract M3HF: Multi-agent Reinforcement Learning from Multi-phase Human Feedback of Mixed Quality MACCA: Offline Multi-agent Reinforcement Learning with Causal Credit Assignment Nover: Incentive Training for Language Models via Verifier-Free Reinforcement Learning ON THE OPTIMIZATION LANDSCAPE OF LOW RANK ADAPTATION METHODS FOR LARGE LANGUAGE MODELS Quantifying the Self-Interest Level of Markov Social Dilemmas RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors Resolving Social Dilemmas with Minimal Reward Transfer - Extended Abstract RuAG: LEARNED-RULE-AUGMENTED GENERATION FOR LARGE LANGUAGE MODELS Towards Communication Efficient Multi-Agent Cooperations: Reinforcement Learning and LLM Towards Communication Efficient Multi-Agent Cooperations: Reinforcement Learning and LLM Will Systems of LLM Agents Lead to Cooperation: An Investigation into a Social Dilemma Aligning Individual and Collective Objectives in Multi-Agent Cooperation An End-to-End Deep Reinforcement Learning Based Modular Task Allocation Framework for Autonomous Mobile Systems A Review of Safe Reinforcement Learning: Methods, Theories and Applications Characterizing Physical Adversarial Attacks on Robot Motion Planners Dual Contrastive Graph-Level Clustering with Multiple Cluster Perspectives Alignment Efficient and scalable reinforcement learning for large-scale network control Explaining an Agent’s Future Beliefs Through Temporally Decomposing Future Reward Estimators Human-Guided Moral Decision Making in Text-based Games Learning the Expected Core of Strictly Convex Stochastic Cooperative Games Learning to Discuss Strategically: A Case Study on One Night Ultimate Werewolf Learning to Discuss Strategically: A Case Study on One Night Ultimate Werewolf Off-Agent Trust Region Policy Optimization PEARL: Zero-shot Cross-task Preference Alignment and Robust Reward Learning for Robotic Manipulation PEARL: Zero-shot Cross-task Preference Alignment and Robust Reward Learning for Robotic Manipulation Policy learning from tutorial books via understanding, rehearsing and introspecting Resolving social dilemmas with minimal reward transfer Safe Reinforcement Learning with Free-form Natural Language Constraints and Pre-Trained Language Models Self-Guiding Exploration for Combinatorial Problems Selfishness Level Induces Cooperation in Sequential Social Dilemmas STAS: Spatial-Temporal Return Decomposition for Multi-agent Reinforcement Learning Tackling Cooperative Incompatibility for Zero-Shot Human-AI Coordination TAPE: Leveraging Agent Topology for Cooperative Multi-Agent Policy Gradient Team-wise Effective Communication in Multi-Agent Reinforcement Learning Temporal Explanations of Deep Reinforcement Learning Agents The Selfishness Level of Social Dilemmas A human-centered safe robot reinforcement learning framework with interactive behaviors A Multi-Agent Framework for Recommendation with Heterogeneous Sources An Efficient End-to-End Training Approach for Zero-Shot Human-AI Coordination ChessGPT: Bridging Policy Learning and Language Modeling Cooperative Multi-Agent Learning in a Complex World: Challenges and Solutions Cooperative Open-ended Learning Framework for Zero-shot Coordination Interpretable Reward Redistribution in Reinforcement Learning: A Causal Approach Invariant Learning via Probability of Sufficient and Necessary Causes Is Nash Equilibrium Approximator Learnable? Leveraging Joint-action Embedding in Multi-agent Reinforcement Learning for Cooperative Games View all publications
9 May 2025 King's scientists receive government funding to enable smooth transition to an AI-powered future Up to £400,000 from the AI Security Institute will fund research exploring how humans and AI can…