Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Moving Out: Physically-grounded Human-AI Collaboration
Xuhui Kang, Sung-Wook Lee, Haolin Liu, Yuyan Wang, Yen-Ling Kuo
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 |
|---|---|---|---|
| add BASS (behavior augmentation, simulation, action selection) | Diffusion Policy · Moving Out Challenge 1 seen behaviors | TCR: 0.3233 -> 0.3503 | Improved |
| add BASS | Diffusion Policy · Moving Out Challenge 1 unseen behaviors | TCR: 0.2563 -> 0.3010 | Improved |
| add BASS | Diffusion Policy · Moving Out Challenge 1, play with human | TCR: 0.3855 -> 0.6512 | Improved |
| add BASS | Diffusion Policy · Moving Out Challenge 1, play with human | Waiting time: 0.4886 -> 0.3364 | Improved |
| add simulation and action selection (vs w/o Simulation) (ablation) | DP + BASS · Moving Out Challenge 1 | TCR: 0.4028 -> 0.5027 | Improved |
| add behavior augmentation (vs w/o Augmentation) (ablation) | DP + BASS · Moving Out Challenge 1 | TCR: 0.4741 -> 0.5027 | Improved |
| add behavior augmentation (vs w/o Augmentation) (ablation) | GRU + BASS · Moving Out Challenge 2 | TCR: 0.3670 -> 0.3414 | Hurt |
| use multi-agent recombination instead of single-agent recombination (ablation) | BASS · Moving Out Challenge 1 | TCR: 0.368 -> 0.403 | Improved |
| use multi-agent action simulation instead of single-agent (ablation) | BASS · Moving Out Challenge 2 | TCR: 0.319 -> 0.420 | Improved |
| replace learned dynamics model with oracle simulator (ablation) | DP + BASS · Moving Out Challenge 2 | NFD: 0.5535 -> 0.6209 | Improved |
| replace learned dynamics model with oracle simulator (ablation) | DP + BASS · Moving Out Challenge 1 | NFD: 0.5733 -> 0.5875 | No clear effect |
| use BASS instead of zero-shot coordination (CoMeDi) | DP/BASS · Moving Out Map 1 unseen proxies | TCR: 0.3768 -> 0.7080 | Improved |
Closest papers by meaning
- Habitat 3.0: A Co-Habitat for Humans, Avatars, and Robots (ICLR 2024 · similarity 0.89)
- Constrained Human-AI Cooperation: An Inclusive Embodied Social Intelligence Challenge (NeurIPS 2024 · similarity 0.86)
- PARTNR: A Benchmark for Planning and Reasoning in Embodied Multi-agent Tasks (ICLR 2025 · similarity 0.86)
- Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models (NeurIPS 2026 · similarity 0.85)
- Beyond Single Stationary Policies: Meta-Task Players as Naturally Superior Collaborators (NeurIPS 2024 · similarity 0.85)
- Adaptive Agents for Mixed-Initiative Human-AI Collaborations (AAAI 2025 · 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.
