Skip to content
Global sound
Embodied AI 101

Embodied AI 101

NEO-VLA: Why Visual Readout Depth Matters for Robot Manipulation

· 13 min

Show notes

A 2B open VLA that feeds pixels directly through a single convolutional layer into a native vision-language backbone, eliminating a separate image encoder, trained on 2.23M episodes and 9.9k hours of robot data. The architectural simplification with strong ablations and fully open code and data makes it a notable contribution to accessible robot foundation models.

Website: share.transistor.fm

Live, last hour: 11,300 songs on 5,979 stations in 146 countries

Discover the world through radio

More ways to discover (9)

Heard something good? Save it, or open it straight away in Spotify.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
Intamixx RadioContinue listening