Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Mixed-Initiative Multiagent Apprenticeship Learning for Human Training of Robot Teams
Esmaeil Seraj, Jerry Yuyang Xiong, Mariah L Schrum, Matthew Gombolay
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 |
|---|---|---|---|
| train MixTURE with MIM differentiable communication instead of heuristic/demonstrated communication (vs MA-GAIL) | MixTURE · FireCommander hard (synthetic) | steps to win (lower better): 80.00 -> 56.49 | Improved |
| MixTURE vs MARL (MA-PPO) on Predator-Prey hard | MixTURE · Predator-Prey hard | steps to win: 47.40 -> 28.73 | Improved |
| MixTURE vs BC+DC (offline BC only, no online loss) on Predator-Capture-Prey medium (ablation) | MixTURE · Predator-Capture-Prey medium | steps to win: 61.54 -> 17.27 | Improved |
| add differentiable communication to non-communicative MA-GAIL (use MixTURE) | MA-GAIL · Predator-Prey medium | steps to win: 60.38 -> 13.15 | Improved |
| add demonstrated communication to MA-GAIL (ablation) | MA-GAIL · Predator-Prey medium | steps to win: 60.38 -> 20.39 | Improved |
| add demonstrated communication to MA-GAIL on Predator-Prey easy (ablation) | MA-GAIL · Predator-Prey easy | steps to win: 17.48 -> 11.68 | Improved |
| add online loss to BC+DC (MixTURE vs BC+DC) on FireCommander easy (ablation) | MixTURE · FireCommander easy | steps to win: 38.95 -> 22.31 | Improved |
| MA-GAIL with demonstrated communication on FireCommander (all difficulties) (ablation) | MA-GAIL · FireCommander | steps to win | No clear effect |
| train on real human data without demonstrated communication (MixTURE) vs MA-GAIL with communication demos | MixTURE · FireCommander human study | NASA-TLX workload: 44.3% reduction | Improved |
| same human-study comparison | MixTURE · FireCommander human study | SUS usability: 46.1% increase | Improved |
| same human-study comparison | MixTURE · FireCommander human study | human game score | Improved |
| same human-study comparison | MixTURE · FireCommander human study | demonstration time per step | Improved |
Closest papers by meaning
- Designs for Enabling Collaboration in Human-Machine Teaming via Interactive and Explainable Systems (NeurIPS 2024 · similarity 0.88)
- Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning (ICLR 2026 · similarity 0.87)
- M³HF: Multi-agent Reinforcement Learning from Multi-phase Human Feedback of Mixed Quality (ICML 2025 · similarity 0.87)
- NestRL: A Nested Training Regime for Mutual Adaptation in Human–AI Teaming (NeurIPS 2026 · similarity 0.87)
- Adaptively Coordinating with Novel Partners via Learned Latent Strategies (NeurIPS 2025 · similarity 0.86)
- Acquiring Diverse Skills using Curriculum Reinforcement Learning with Mixture of Experts (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.
