Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Improving Human-AI Coordination through Online Adversarial Training and Generative Models
Paresh R. Chaudhary, Yancheng Liang, Daphne Chen, Simon Shaolei Du, Natasha Jaques
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| use a regret objective instead of minimax for the adversary (ablation) | GOAT · Overcooked Multi-Strategy Counter, held-out BC human proxy | cooperation score | Improved |
| train the cooperator against a GOAT adversary (generative model plus regret) instead of GAMMA | Cooperator agent · Overcooked Multi-Strategy Counter, real humans | team score: 38% improvement over the previous state of the art | Improved |
| train the cooperator against a GOAT adversary instead of GAMMA | Cooperator agent · Overcooked Counter Circuit, real humans | team score: 3% improvement over GAMMA | Improved |
Closest papers by meaning
- Learning to Cooperate with Humans using Generative Agents (NeurIPS 2024 · similarity 0.89)
- Automated Red Teaming with GOAT: the Generative Offensive Agent Tester (ICML 2025 · similarity 0.88)
- Robust and Diverse Multi-Agent Learning via Rational Policy Gradient (NeurIPS 2025 · similarity 0.86)
- Adaptively Coordinating with Novel Partners via Learned Latent Strategies (NeurIPS 2025 · similarity 0.85)
- ROTATE: Regret-driven Open-ended Training for Ad Hoc Teamwork (NeurIPS 2026 · similarity 0.85)
- ALSO: Adversarial Online Strategy Optimization for Social Agents (ICML 2026 · similarity 0.85)
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.
