AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

VER: Vision Expert Transformer for Robot Learning via Foundation Distillation and Dynamic Routing

Yixiao Wang, Mingxiao Huo, Zhixuan Liang, Yushi Du, Lingfeng Sun, Haotian Lin, Jinghuan Shang, Chensheng Peng, Mohit Bansal, Mingyu Ding, Masayoshi Tomizuka

ICLR 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 2 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
use VER-B instead of Theia-Bpolicy head on frozen encoder · 11 manipulation taskssuccess rate (%): 67.1 -> 74.7Improved
use VER-T instead of Theia-Tflow-matching head · cross->binsuccess rate: 0.65 -> 0.95Improved
use VER-T instead of Theia-Tdiffusion head · Real-world poursuccess rate: 0.45 -> 0.90Improved
use VER instead of fine-tuned GR00T N1.5flow-matching head · cube->cupsuccess rate: 0.73 -> 0.75Improved
add curriculum Top-K annealing to patchwise expert routing (ablation)VER · Relocatesuccess rate (%): 47.6 -> 56.4Improved
add curriculum Top-K annealing to patchwise expert routing (ablation)VER · Pensuccess rate (%): 78.0 -> 80.8Improved
use patchwise expert routing instead of a single frozen teacher router (ablation)VER · Relocatesuccess rate (%): 38.4 -> 47.6Improved
use patchwise instead of layerwise teacher routing (ablation)VER · Relocatesuccess rate (%): 36.4 -> 47.6Improved
increase Top-K from 1 to 3 (ablation)VER-Tiny · Relocate+Pen averagesuccess rate (%): 60.0 -> 68.0Improved
add one train-from-scratch expert to distilled experts (ablation)VER · Relocate+Pen averagesuccess rate (%): 72.0 -> 78.7Improved
increase CTA schedule S from 0 to 60 (ablation)VER · pen-v0success rate (%): 77.3 -> 81.3Improved
scale experts from (K,L)=(1,3) to (2,6) (ablation)VER-Tiny · pen-v0success rate (%): 44.0 -> 69.6Improved

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