Sharpa's SALA aligns latent actions across human and robot video for dual-axis scaling
Jiachun Jin · hf · 2026-11-17
A paper from Sharpa Robotics and the SJTU-Sharpa joint lab introduces SALA. Most latent action models learn latent actions via video reconstruction, so representations cluster by visual appearance rather than underlying action, making it hard to align human motions with robot counterparts. SALA uses language-based semantic similarity to align latent actions from egocentric human and robot videos into a shared latent action space that supervises embodied foundation model training, enabling dual-axis scaling and improving robot task performance while avoiding costly explicit pixel prediction.
More from Embodied
- Italian startup Mosaico raises €7.5M to fix robotics data for physical AI — lukas_m_ziegler · 2026-11-17
- Video made entirely via Neuralink implant by ALS patient, retweeted by Musk — elonmusk · 2026-11-17
- Skild AI open-sources Steel physics solver: up to 13x faster than PhysX with 3x less GPU memory — deepakpathak · 2026-11-17
- XREAL Aura review: the spatial computer Apple should have made — Scobleizer · 2026-11-17
- ScaleVLA scales VLA models to 552B parameters: 10 demos beat 2B backbone trained on 50 — GalaxyGeneralRobotics · 2026-11-17
- Amazon's next delivery runner has four legs: RIVR robot joins the last-mile fleet — TinfoilTricorn · 2026-11-17