AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

Haozhan Li, Yuxin Zuo, Jiale Yu, Yuhao Zhang, Yang Zhaohui, Kaiyan Zhang, Xuekai Zhu, Yuchen Zhang, Tianxing Chen, Ganqu Cui, Dehui Wang, Dingxiang Luo and 9 more

ICLR 2026 · Reinforcement learning › RL from feedback · approach: RL for LLM reasoning

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
apply SimpleVLA-RL on top of full-data SFTOpenVLA-OFT · LIBERO average (Full-Trajectory SFT)success rate (%): 91.0 -> 99.1Improved
apply SimpleVLA-RL on top of one-demonstration-per-task SFTOpenVLA-OFT · LIBERO-Long (One-Trajectory SFT)success rate (%): 17.3 -> 91.7Improved
apply SimpleVLA-RL on top of SFT for dual-arm tasksOpenVLA-OFT · RoboTwin2.0 (12 tasks overall average)success rate (%): 38.3 -> 68.8Improved
apply SimpleVLA-RL on SFT model trained only in simulation, evaluated on real robotOpenVLA-OFT · Real-world sim2real (4 tasks average)success rate (%): 17.5 -> 38.5Improved
use RL instead of SFT on 9 seen tasks, evaluating on held-out unseen tasks (ablation)OpenVLA-OFT (One-Trajectory SFT base) · LIBERO Spatial/Object/Goal unseen taskssuccess rate on unseen task: Spatial Unseen Task 1: 43.3 -> 71.8Improved
add exploration enhancements: dynamic sampling, clip higher (0.28), temperature 1.6 (ablation)SimpleVLA-RL · LIBEROsuccess rate: 10-15% consistent improvementImproved
apply RL to a base model with no SFT trajectories (ablation)OpenVLA-OFT · RoboTwin2.0 5 tasks, 0 trajectories SFTsuccess rate (%): 0 -> 0No clear effect
increase SFT trajectories from 100 to 1000 before RL (ablation)SimpleVLA-RL · RoboTwin2.0 5 tasks averagesuccess rate after RL (%): 25.4 -> 50.4Improved
apply RL when initial success rate is very low (100 trajectories SFT) (ablation)SimpleVLA-RL · RoboTwin2.0 Pick Dual Bottlessuccess rate (%): 1.2 -> 4.3 (vs 29.7 -> 68.3 with 1000 trajectories)Mixed
remove KL regularization and reference modelSimpleVLA-RL · trainingmemory and explorationImproved

Later papers that compared with this method

1 later accepted paper reports SimpleVLA-RL as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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