Research map / Multi-agent systems and games / Cooperative multi-agent RL
A Cordial Sync: Going Beyond Marginal Policies for Multi-Agent Embodied Tasks
Unnat Jain, Luca Weihs, Eric Kolve, Ali Farhadi, Svetlana Lazebnik, Aniruddha Kembhavi, Alexander Schwing
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 |
|---|---|---|---|
| increase mixture components from 1 to 2 (ablation) | SYNC · FurnMove | Success rate: 0.328 -> 0.502 | Improved |
| increase mixture components from 2 to 4 (ablation) | SYNC · FurnMove | Success rate: 0.502 -> 0.569 | Improved |
| increase mixture components from 4 to 13 (ablation) | SYNC · FurnMove | Success rate: 0.569 -> 0.587 | Improved |
| increase mixture components from 2 to 13 (ablation) | SYNC · FurnMove | TVD: 0.206 -> 0.474 | Improved |
| add coordination loss (ablation) | SYNC · FurnMove | Success rate: 0.488 -> 0.587 | Improved |
| add coordination loss (ablation) | SYNC · FurnMove | Invalid action probability: 0.47 -> 0.31 | Improved |
| add coordination loss (ablation) | Central · FurnMove | Success rate: 0.609 -> 0.648 | Improved |
| add coordination loss (ablation) | Marginal · FurnMove | Success rate: 0.328 -> 0.099 | Hurt |
| replace Marginal policy with SYNC | Marginal · FurnMove | Success rate: 0.328 -> 0.587 | Improved |
| replace Marginal policy with SYNC | Marginal · FurnMove-Grid | Success rate: 0.694 -> 0.762 | Improved |
| replace Marginal policy with SYNC | Marginal · FurnLift | Success rate: 0.449 -> 0.423 | No clear effect |
| replace Marginal policy with SYNC | Marginal · FurnLift | Failed pickups: 8.933 -> 4.873 | Improved |
Closest papers by meaning
- Collaborative Tree Search for Enhancing Embodied Multi-Agent Collaboration (CVPR 2025 · similarity 0.89)
- Mosaic: Runtime-Efficient Multi-Agent Embodied Planning (ICML 2026 · similarity 0.88)
- LLawCo: Learning Laws of Cooperation for Modeling Embodied Multi-Agent Behavior (ICML 2026 · similarity 0.88)
- Beyond Single Stationary Policies: Meta-Task Players as Naturally Superior Collaborators (NeurIPS 2024 · similarity 0.87)
- CaPo: Cooperative Plan Optimization for Efficient Embodied Multi-Agent Cooperation (ICLR 2025 · similarity 0.87)
- COMBO: Compositional World Models for Embodied Multi-Agent Cooperation (ICLR 2025 · similarity 0.87)
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.
