A 2M-parameter MAPF-GPT model fine-tuned with delta-driven data generation matches or exceeds much larger models on MAPF benchmarks and scales to over a million agents.
Prioritized multi- agent path finding for differential drive robots,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Advancing Learnable Multi-Agent Pathfinding Solvers with Active Fine-Tuning
A 2M-parameter MAPF-GPT model fine-tuned with delta-driven data generation matches or exceeds much larger models on MAPF benchmarks and scales to over a million agents.