Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Emergent Coordination Through Competition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add population-based training with evolution (ablation) | feedforward agents in 2v2 MuJoCo soccer · 2v2 MuJoCo soccer | agent performance | Helped |
| add dense shaping rewards (ablation) | evolving feedforward agents · 2v2 MuJoCo soccer | early learning performance against the dummy evaluator | Helped |
| add a recurrent action-value function (ablation) | evolving agents with dense shaping rewards · 2v2 MuJoCo soccer | training steps to reach weighted expected goal difference 0: 22B vs 35B | Hurt |
| use a recurrent policy with a recurrent action-value function (ablation) | evolving agents with dense shaping rewards · 2v2 MuJoCo soccer | performance against Nash-averaged evaluators | Hurt |
| add separate reward-channel action-value estimates and individual discount factors (ablation) | recurrent-critic agents with dense shaping rewards · 2v2 MuJoCo soccer | training steps to reach weighted expected goal difference 0.2: 15B vs 35B | Helped |
| add separate reward channels (ablation) | agents trained with a recurrent policy and critic · 2v2 MuJoCo soccer, round-robin tournament | Elo | Helped |
| add a recurrent actor (ablation) | agents with a recurrent critic, shaping rewards, and separate reward channels · 2v2 MuJoCo soccer, round-robin tournament | Elo | Helped |
| automatically evolve shaping reward weights using match results (ablation) | population-trained soccer agents · 2v2 MuJoCo soccer | long-horizon team behavior | Helped |
Closest papers by meaning
- Mimicking To Dominate: Imitation Learning Strategies for Success in Multiagent Games (NeurIPS 2024 · similarity 0.88)
- Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination (ICML 2025 · similarity 0.88)
- Multi-agent cooperation through learning-aware policy gradients (ICLR 2025 · similarity 0.87)
- Foundations of Multi-Agent Learning in Dynamic Environments: Where Reinforcement Learning Meets Strategic Decision-Making (AAAI 2025 · similarity 0.86)
- Stability of Multi-Agent Learning in Competitive Networks: Delaying the Onset of Chaos (AAAI 2024 · similarity 0.86)
- Efficient Adaptation in Mixed-Motive Environments via Hierarchical Opponent Modeling and Planning (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.