AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

RDT2: Exploring the Scaling Limit of UMI Data Towards Zero-Shot Cross-Embodiment Generalization

Songming Liu, Bangguo Li, Kai Ma, Lingxuan Wu, Hengkai Tan, Xiao Ouyang, Hang Su, Jun Zhu

ICML 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add AR (RVQ cross-entropy) pretraining stage before diffusion (ablation)RDT2 · Training lossconvergence speed and final lossImproved
replace FAST tokenizer with RVQ tokenizer (ablation)RDT2 action discretization · Discretization error vs tokenstokens needed at equal error: saves up to about two-thirds of the tokensImproved
use uniform binning instead of RVQ (ablation)RDT2 action discretization · Discretization error vs tokenstoken efficiency: lowest error but far more tokensHurt
add one-step diffusion distillation (Stage 3) (ablation)RDT2-FM · Inference frequencyHz: 16.0 Hz -> 23.0 HzImproved
add flow-matching action expert (Stage 2) over RDT2-VQ (ablation)RDT2-VQ · Inference frequencyHz: 2.7 Hz -> 16.0 HzImproved
flow-matching expert vs RVQ autoregressive head (ablation)RDT2 · Zero-shot tasks (unseen embodiment)success rateNo clear effect
scale model size and data jointly (ablation)RDT2-VQ · Training loss (one epoch)training lossImproved
fine-tune RDT2 instead of pi0.5pi0.5 · Cloth foldingsuccess rate (%): 36 -> 77Improved
fine-tune RDT2 instead of pi0.5pi0.5 · Unzippingsuccess rate (%): 13 -> 45Improved
fine-tune RDT2 instead of pi0.5pi0.5 · Button pressingreaction time gap (ms): +323 -> +97Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy