SSAD talk breaks down building a driving VLA in 4 steps: model, data, action head, finetuning
abursuc · x · 2026-09-17
At the SSAD 2026 workshop, a talk walked through building a VLA (vision-language-action) model for driving in four steps: model selection, data curation, action-head design, and multi-stage finetuning.
It opened with a brief history of the autonomous driving stack and where the field is headed next.
Related event: SSAD Talk Details Building Compact Driving VLA Models(3 posts)→
More from Embodied
- Ex-Tsinghua/BIGAI team's generalist dexterity policy beats Astra on RoboDojo — chris_j_paxton · 2026-09-17
- GPT-Policy: In-Context Robot Learning with VLM Agents, No Gradient Updates — Dongzhou Cheng · 2026-09-17
- World Labs' Atlas Scans by Generative Guessing; NeRF Creator Admits Productization Is Hard — cen6wkf · 2026-09-17
- CXMT's LPDDR5X lands in flagship phone as Nubia ships $885 Doubao AI handset — pstAsiatech · 2026-09-17
- Key open challenges for VLAs: language, evaluation, deployment, causal reasoning — abursuc · 2026-09-17
- LADA: latent actions imitate language from few observation-language pairs — abursuc · 2026-09-17