AnySlot: Goal-Conditioned Vision-Language-Action Policies for Zero-Shot Slot-Level Placement

Zhaofeng Hu *, Sifan Zhou *†, Qinbo Zhang, Rongtao Xu, Qi Su, Ci-Jyun Liang † · 2026

Read AnySlot: Goal-Conditioned Vision-Language-Action Policies for Zero-Shot Slot-Level Placement by Zhaofeng Hu *, Sifan Zhou *†, Qinbo Zhang, Rongtao Xu, Qi Su, Ci-Jyun Liang † (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-10432.txt.

Description

Hu and six colleagues teach robots to drop items into exact slots from spoken orders. Flat vision language action nets fumble fine placement under layered word commands. Their AnySlot system first paints a visual mark on the target slot, then steers by that mark. Splitting slot choice from low level control lifts both word sense and spatial care. A new nine part test set, SlotBench, shows clear wins over flat and modular rivals in fresh tasks.

Catalog record

CREATORZhaofeng Hu *, Sifan Zhou *†, Qinbo Zhang, Rongtao Xu, Qi Su, Ci-Jyun Liang †
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39376
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.10432

Browse the full library · About the Tower of Babel