ICML 2025 accepted papers on Multi-agent systems and games
97 of the 3,342 papers accepted at ICML 2025 (2.9%) are on Multi-agent systems and games, compared with 2.0% at ICML 2024. They fall into 4 topics.
See ICML 2025 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
LLM agents · 40 papers
- Agent Workflow Memory (10 citations)
- AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML (5 citations)
- ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Tasks (4 citations)
- Agent-as-a-Judge: Evaluate Agents with Agents (3 citations)
- Cradle: Empowering Foundation Agents towards General Computer Control (3 citations)
- CVE-Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World Web Application Vulnerabilities (2 citations)
- RE-Bench: Evaluating Frontier AI R&D Capabilities of Language Model Agents against Human Experts (2 citations)
- Windows Agent Arena: Evaluating Multi-Modal OS Agents at Scale (2 citations)
- G-Designer: Architecting Multi-agent Communication Topologies via Graph Neural Networks (1 citation)
- On the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents (1 citation)
- Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks (1 citation)
- ShieldAgent: Shielding Agents via Verifiable Safety Policy Reasoning (1 citation)
and 28 more in the interactive map.
Human-AI and goal-conditioned · 24 papers
- Position: Truly Self-Improving Agents Require Intrinsic Metacognitive Learning (7 citations)
- Scaling Laws for Pre-training Agents and World Models (2 citations)
- MCU: An Evaluation Framework for Open-Ended Game Agents (1 citation)
- A Theoretical Study of (Hyper) Self-Attention through the Lens of Interactions: Representation, Training, Generalization
- Accurate and Efficient World Modeling with Masked Latent Transformers
- Ad-Hoc Human-AI Coordination Challenge
- Closed-Loop Long-Horizon Robotic Planning via Equilibrium Sequence Modeling
- Craftium: Bridging Flexibility and Efficiency for Rich 3D Single- and Multi-Agent Environments
- Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination
- Hybrid Quantum-Classical Multi-Agent Pathfinding
- KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems
- NBDI: A Simple and Effective Termination Condition for Skill Extraction from Task-Agnostic Demonstrations
and 12 more in the interactive map.
Cooperative multi-agent RL · 18 papers
- R3DM: Enabling Role Discovery and Diversity Through Dynamics Models in Multi-agent Reinforcement Learning (1 citation)
- Distributionally Robust Multi-Agent Reinforcement Learning for Dynamic Chute Mapping
- Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration
- Learning Mean Field Control on Sparse Graphs
- Learning Progress Driven Multi-Agent Curriculum
- O-MAPL: Offline Multi-agent Preference Learning
- Sable: a Performant, Efficient and Scalable Sequence Model for MARL
- Wolfpack Adversarial Attack for Robust Multi-Agent Reinforcement Learning
- Agent-Centric Actor-Critic for Asynchronous Multi-Agent Reinforcement Learning
- AI for Global Climate Cooperation: Modeling Global Climate Negotiations, Agreements, and Long-Term Cooperation in RICE-N
- Finite-Time Global Optimality Convergence in Deep Neural Actor-Critic Methods for Decentralized Multi-Agent Reinforcement Learning
- GradPS: Resolving Futile Neurons in Parameter Sharing Network for Multi-Agent Reinforcement Learning
and 6 more in the interactive map.
Games and equilibria · 15 papers
- Self-Play $Q$-Learners Can Provably Collude in the Iterated Prisoner's Dilemma (2 citations)
- From Debate to Equilibrium: Belief‑Driven Multi‑Agent LLM Reasoning via Bayesian Nash Equilibrium (1 citation)
- Incentivize without Bonus: Provably Efficient Model-based Online Multi-agent RL for Markov Games (1 citation)
- Observation Interference in Partially Observable Assistance Games (1 citation)
- Adaptive Sample Sharing for Multi Agent Linear Bandits
- Breaking the Curse of Multiagency in Robust Multi-Agent Reinforcement Learning
- Contract Design Under Approximate Best Responses
- Convex Markov Games: A New Frontier for Multi-Agent Reinforcement Learning
- Counterfactual Effect Decomposition in Multi-Agent Sequential Decision Making
- Learning to Stop: Deep Learning for Mean Field Optimal Stopping
- PAC Learning with Improvements
- Prediction-Aware Learning in Multi-Agent Systems
and 3 more in the interactive map.