Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Truncated Counterfactual Learning for Anytime Multi-Agent Path Finding
Thomy Phan, Shao-Hung Chan, 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 |
|---|---|---|---|
| reset MABUC parameters when top-K set changes (non-stationarity) (ablation) | TACKLE · Warehouse and City, 800 agents | sum of delays and AUC | Improved |
| use counterfactual TACKLE instead of interventional ADDRESS | MAPF-LNS seed selection · Warehouse and City, 800 agents | sum of delays | Improved |
| use TACKLE instead of observational Roulette Wheel or Tabu List | delay-based MAPF-LNS · Warehouse and City, 800 agents | sum of delays | Improved |
| use TACKLE instead of Roulette Wheel at m=1000 | delay-based MAPF-LNS · Warehouse | cost: at least 76% cost improvement | Improved |
| use TACKLE instead of Roulette Wheel at m=1000 | delay-based MAPF-LNS · City | cost: at least 60% cost improvement | Improved |
| use TACKLE instead of LaCAM* | anytime MAPF · Warehouse | sum of delays: 70, 107.8 -> 411.1 | Improved |
| increase truncation K from 32 to 128 or more (ablation) | ADDRESS · Warehouse and City, 800 agents | sum of delays | Hurt |
| increase truncation K to large values (ablation) | TACKLE · Warehouse and City, 800 agents | sum of delays | Hurt |
| use TACKLE instead of baselines on small Random map | MAPF-LNS · Random-32-32-20 | sum of delays | No clear effect |
Closest papers by meaning
- Anytime Multi-Agent Path Finding with an Adaptive Delay-Based Heuristic (AAAI 2025 · similarity 0.93)
- Adaptive Anytime Multi-Agent Path Finding Using Bandit-Based Large Neighborhood Search (AAAI 2024 · similarity 0.93)
- Decentralized Monte Carlo Tree Search for Partially Observable Multi-Agent Pathfinding (AAAI 2024 · similarity 0.90)
- Learn to Follow: Decentralized Lifelong Multi-Agent Pathfinding via Planning and Learning (AAAI 2024 · similarity 0.90)
- LNS2+RL: Combining Multi-agent Reinforcement Learning with Large Neighborhood Search in Multi-agent Path Finding (AAAI 2025 · similarity 0.89)
- MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at Scale (AAAI 2025 · similarity 0.89)
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.
