Skip to main content
← Back to news

NEO-VLA: the encoder-free vision-action policy, the model that watches the world in pixels

Research
NEO-VLA (MMLab, NTU) is an encoder-free vision-language-action framework: pixel images evolve with language across depths and a flow-matching action expert uses representations from multiple layers. Native architectures outperform encoder-based ones.

Get the next local-AI briefing

Once a week, the top selection from iaLocale: articles, news and resources on local AI, and more.

After signing up you receive a confirmation email (double opt-in). Without confirmation we send nothing. No spam, we promise.