A minimax-Bayes choice of teammate-training distribution improves worst-case robustness in ad hoc teamwork, with the strongest empirical gains on Overcooked and Melting Pot.
Title resolution pending
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
-
A Minimax Approach to Ad Hoc Teamwork
A minimax-Bayes choice of teammate-training distribution improves worst-case robustness in ad hoc teamwork, with the strongest empirical gains on Overcooked and Melting Pot.