--- format: "aidr-story-markdown/v1" id: "bf847f4610885a0948262844a710e2fe7ec48b26f6f3df4529ea05f2013770ab" canonical_url: "https://aidr.today/bf847f46?lang=en" title: "From Pixels to Actions - Towards Native Vision-Language-Action Policies" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-11-05T00:00:00.000Z" category: "Research" topics: ["agent"] source_urls: ["https://huggingface.co/papers/2611.06918"] summary: "Researchers propose NEO-VLA, an encoder-free vision-language-action framework that tokenizes images with lightweight patch convolutional layers and evolves them jointly with language throughout the backbone. The approach allows a flow-matching action expert to draw from instruction-conditioned visual representations at different depths, and native architectures reportedly outperform encoder-based counterparts under controlled comparisons. Layer-wise probing reportedly shows that action-critical visual signals often peak well before the final visual representation." --- # From Pixels to Actions \- Towards Native Vision\-Language\-Action Policies > [Open the canonical story]() **Published:** 2026-11-05T00:00:00.000Z **Category:** Research **Topics:** agent ## Summary Researchers propose NEO\-VLA, an encoder\-free vision\-language\-action framework that tokenizes images with lightweight patch convolutional layers and evolves them jointly with language throughout the backbone\. The approach allows a flow\-matching action expert to draw from instruction\-conditioned visual representations at different depths, and native architectures reportedly outperform encoder\-based counterparts under controlled comparisons\. Layer\-wise probing reportedly shows that action\-critical visual signals often peak well before the final visual representation\. ## Sources - [Story source]()