Premover enables VLA policies to act on partial instructions by precomputing focus maps from intermediate backbone layers, reducing wall-clock time 13.6 percent on LIBERO while preserving 95 percent success rate.
Zhao, Vikash Kumar, Sergey Levine, and Chelsea Finn
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Premover: Fast Vision-Language-Action Control by Acting Before Instructions Are Complete
Premover enables VLA policies to act on partial instructions by precomputing focus maps from intermediate backbone layers, reducing wall-clock time 13.6 percent on LIBERO while preserving 95 percent success rate.