DLM unifies offline multi-agent sequential decision making by modeling it as dialogue-style sequence prediction with two-stage LLM training, outperforming baselines and showing zero-shot generalization.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DLM: Unified Decision Language Models for Offline Multi-Agent Sequential Decision Making
DLM unifies offline multi-agent sequential decision making by modeling it as dialogue-style sequence prediction with two-stage LLM training, outperforming baselines and showing zero-shot generalization.