AdaCred trains decision transformers with learned GumbelSigmoid token masks and an efficiency loss, claiming improved offline RL and imitation learning with shorter, pruned sequences.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
AdaCred: Adaptive Causal Decision Transformers with Feature Crediting
AdaCred trains decision transformers with learned GumbelSigmoid token masks and an efficiency loss, claiming improved offline RL and imitation learning with shorter, pruned sequences.