NAT: streaming action tokenizer closes latency gap with flow action heads, trained on 49k hours
YuXiang_IRVL · x · 2026-11-22
Researchers introduce NAT, a streaming action tokenizer for VLAs that closes the inference latency gap with flow action heads. Trained on 49k hours of robot data, it can zero-shot tokenize unseen robotic embodiments.
- Targets the slow inference of flow-based action heads in VLA models
- Zero-shot generalization to new robot embodiments
More from Embodied
- Dallas riders ditch rival robotaxis after trying Tesla's Cybercab — mitchdeg · 2026-11-22
- RoboDojo sim benchmark audit finds loose joints, ignored material properties — AIFlow_ML · 2026-11-22
- LeCun coauthors prove standard world models fail at counterfactuals, fix needs no retraining — alex_verem · 2026-11-22
- Heavy-duty robot shows off gym-bro strength in new demo — LinusEkenstam · 2026-11-22
- Slush anecdote: 50 people pitched data/ML platforms, zero sold robotics hardware — MarvinTBaumann · 2026-11-22
- Caltech's ASTRA robot autonomously builds 1.4m antenna in orbit, a space first — mattbeane · 2026-11-22