A task-conditioned discrete representation objective with a Wasserstein regularizer, claiming an accuracy versus sample-complexity trade-off, with applications to RL state abstraction and domain generalization.
Uniform deviation bounds for k-means clustering
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Task-Driven Discrete Representation Learning
A task-conditioned discrete representation objective with a Wasserstein regularizer, claiming an accuracy versus sample-complexity trade-off, with applications to RL state abstraction and domain generalization.