Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Spatially Grouped Curriculum Learning for Multi-Agent Path Finding
Thomy Phan, Sven Koenig
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 grouping mask M to attention critic (ablation) | PARCEL · Random map, 64 agents (training) | training progress | Improved |
| use self-attention critic instead of linear grouping (VAST) (ablation) | PARCEL · Random map, 64 agents (training) | training progress | Improved |
| use PARCEL instead of independent PPO (ablation) | PARCEL · Random map, 64 agents (training) | training progress | Improved |
| use random groups (REFIL) instead of spatial groups (ablation) | PARCEL · Random map, 64 agents (training) | training progress | Hurt |
| use PARCEL attention critic instead of SCRIMP attention actor and CACTUS centralized critic | PARCEL · Stag Hunt | stag rate | Improved |
| use PARCEL instead of CACTUS | RL-based MAPF · Random and Warehouse, 64 agents (training) | completion rate | Improved |
| use PARCEL instead of CACTUS with 16 agents | RL-based MAPF · Random and Warehouse, 16 agents (training) | completion rate | No clear effect |
| use PARCEL instead of PRIMAL, SCRIMP, CostTracer | RL-based MAPF · Den520d and Paris_1_256 (test, unseen maps) | completion rate and travel time | Improved |
| use PARCEL instead of LaCAM* | MAPF solver · Den520d and Paris_1_256 (test) | completion rate and travel time | Hurt |
| use PRIMAL in small 3x3 MAPF task | PRIMAL · 3x3 grid MAPF (2 agents) | completion rate | Hurt |
Closest papers by meaning
- LNS2+RL: Combining Multi-agent Reinforcement Learning with Large Neighborhood Search in Multi-agent Path Finding (AAAI 2025 · similarity 0.91)
- MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at Scale (AAAI 2025 · similarity 0.90)
- Learn to Follow: Decentralized Lifelong Multi-Agent Pathfinding via Planning and Learning (AAAI 2024 · similarity 0.89)
- Decentralized Monte Carlo Tree Search for Partially Observable Multi-Agent Pathfinding (AAAI 2024 · similarity 0.88)
- Curriculum Design for Trajectory-Constrained Agent: Compressing Chain-of-Thought Tokens in LLMs (NeurIPS 2025 · similarity 0.86)
- POGEMA: A Benchmark Platform for Cooperative Multi-Agent Pathfinding (ICLR 2025 · 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.
