NEO-VLA: native VLA policies drop the separate vision encoder, tested on RoboTwin C2R
liuziwei7 · x · 2026-11-06
- LMMs-Lab announces NEO-VLA, a native vision-language-action policy that processes visual and language tokens jointly instead of relying on a separate pretrained vision encoder.
- Action policies read rich visual representations across backbone layers.
- The team highlights native pretraining, intermediate visual features, and continued understanding during post-training, with promising results on RoboTwin C2R.
Related event: NEO-VLA: Native Vision-Language-Action Policy Without Vision Encoder(2 posts)→
More from Embodied
- Why hasn't anyone hooked an AI agent to a swarm of remote-controlled cyborg cockroaches? — Sauers_ · 2026-11-06
- Peking University-led team open-sources 50K-hour recipe for embodied foundation models — DAGroup-PKU · 2026-11-06
- Physical AI meetup in India: 4 robotics founders on real factory deployments and eval loops — carrycooldude · 2026-11-06
- Northwestern turns ordinary smartwatches into hand and gesture sensors with TransfHAR and electrode-based system — realkaranahuja · 2026-11-06
- Industrial arms earn 42% of robot revenue but only 15% of its value: analysis — ai · 2026-11-06
- Hacker reverse-engineers Apple's unreleased touchscreen MacBook HID APIs — yoimnotkesku · 2026-11-06