ICLR 2025 accepted papers on Multi-agent systems and games
119 of the 3,703 papers accepted at ICLR 2025 (3.2%) are on Multi-agent systems and games, compared with 2.3% at ICLR 2024. They fall into 4 topics.
See ICLR 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 · 59 papers
- BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments (16 citations)
- OpenHands: An Open Platform for AI Software Developers as Generalist Agents (15 citations)
- Automated Design of Agentic Systems (8 citations)
- MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering (7 citations)
- Scaling Large Language Model-based Multi-Agent Collaboration (7 citations)
- Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents (6 citations)
- AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents (5 citations)
- AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents (3 citations)
- AgentSquare: Automatic LLM Agent Search in Modular Design Space (3 citations)
- Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems (3 citations)
- Hypothetical Minds: Scaffolding Theory of Mind for Multi-Agent Tasks with Large Language Models (3 citations)
- Internet of Agents: Weaving a Web of Heterogeneous Agents for Collaborative Intelligence (3 citations)
and 47 more in the interactive map.
Human-AI and goal-conditioned · 28 papers
- OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Code (3 citations)
- VisualPredicator: Learning Abstract World Models with Neuro-Symbolic Predicates for Robot Planning (3 citations)
- Dream to Manipulate: Compositional World Models Empowering Robot Imitation Learning with Imagination (2 citations)
- MetaUrban: An Embodied AI Simulation Platform for Urban Micromobility (2 citations)
- Simulating Human-like Daily Activities with Desire-driven Autonomy (2 citations)
- ThinkBot: Embodied Instruction Following with Thought Chain Reasoning (1 citation)
- ADAM: An Embodied Causal Agent in Open-World Environments
- Agent-to-Sim: Learning Interactive Behavior Models from Casual Longitudinal Videos
- CtD: Composition through Decomposition in Emergent Communication
- DataEnvGym: Data Generation Agents in Teacher Environments with Student Feedback
- Efficient Active Imitation Learning with Random Network Distillation
- Efficient Exploration and Discriminative World Model Learning with an Object-Centric Abstraction
and 16 more in the interactive map.
Cooperative multi-agent RL · 27 papers
- POGEMA: A Benchmark Platform for Cooperative Multi-Agent Pathfinding (6 citations)
- COMBO: Compositional World Models for Embodied Multi-Agent Cooperation (2 citations)
- IntersectionZoo: Eco-driving for Benchmarking Multi-Agent Contextual Reinforcement Learning (1 citation)
- Learning to Steer Markovian Agents under Model Uncertainty (1 citation)
- A Generalist Hanabi Agent
- Accelerating Task Generalisation with Multi-Level Skill Hierarchies
- ComaDICE: Offline Cooperative Multi-Agent Reinforcement Learning with Stationary Distribution Shift Regularization
- Discrete GCBF Proximal Policy Optimization for Multi-agent Safe Optimal Control
- eQMARL: Entangled Quantum Multi-Agent Reinforcement Learning for Distributed Cooperation over Quantum Channels
- Expected Return Symmetries
- Exploiting Structure in Offline Multi-Agent RL: The Benefits of Low Interaction Rank
- Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning
and 15 more in the interactive map.
Games and equilibria · 5 papers
- Advantage Alignment Algorithms (1 citation)
- A Policy-Gradient Approach to Solving Imperfect-Information Games with Best-Iterate Convergence
- Strategic Classification With Externalities
- Single-agent Poisoning Attacks Suffice to Ruin Multi-Agent Learning
- Tractable Multi-Agent Reinforcement Learning through Behavioral Economics