Skip to main content
Search

Amit Jadhav

1 related articles

Subscribe to updates
  1. arXiv

    Moto's action vocabulary no longer hinges on the seed: macaque premotor geometry is imposed on its discrete codes and still aligns on held-out shapes

    In robot models such as Moto that learn discrete action vocabularies from unlabeled video, the training objective fixes frame reconstruction but not which motion receives which code: two independently trained tokenizers assign nearly unrelated codes, and evaluating one tokenizer in bf16 reassigns 74% of its tokens. The authors impose the representational geometry of macaque ventral premotor cortex (PMv) during shape drawing as a prior, using dual-stream training with a straight-through similarity loss to reshape the code assignment; on five shape categories the loss never saw, the PMv-trained tokenizer aligns with PMv better than the released tokenizer and two no-prior runs, and better than the session-averaged 95th percentile of 100 random kinematic encoders, 0.42.

You’re all caught up