FlexLAM trains variable-length latent actions with nested dropout to match or exceed fixed-capacity LAMs at every token budget under scarce-label supervision without new architectures or losses.
Track2Act: Predicting Point Tracks from Internet Videos Enables Generalizable Robot Manipulation , ISBN=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FlexLAM: Resolving the Bottleneck Trade-off in Latent Action Learning
FlexLAM trains variable-length latent actions with nested dropout to match or exceed fixed-capacity LAMs at every token budget under scarce-label supervision without new architectures or losses.