arXiv AI October 12, 2026 By Xule Gao, Xi Wang, Zehua Zang, Rui Wang, Changwen Zheng, Chuxiong Sun

OGPO: Offline Goal-conditioned Policy Optimization for Recoverable Vision-Language-Action Models

OGPO is an Offline Goal‑conditioned Policy Optimization framework designed to improve Vision‑Language‑Action (VLA) models for robotic manipulation. It replaces traditional step‑wise action imitation with adaptive relabeling of trajectories using semantic key states and goal‑conditioned rewards, enabling policies to reach task‑progress states even from off‑demonstration states. Experiments on LIBERO, MetaWorld, LIBERO‑Plus, and the proposed LIBERO‑ReAct benchmark show that OGPO consistently outperforms standard supervised fine‑tuning in terms of robustness and recoverability under distribution shifts.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI

Jul 1

Z-1: Efficient Reinforcement Learning for Vision-Language-Action Models

arXiv:2606. 31846v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models offer a promising framework for robotic manipulation by connecting language instructions, visual observations, and continuous control.

By Lang Cao, Renhong Chen, Luyi Li, Peng Wang, Mofan Peng, Yitong Li

reinforcement-learning robotics fine-tuning multimodal benchmarks

arXiv AI

Oct 2

Rewind-IL: Online Failure Detection and State Respawning for Imitation Learning

Rewind-IL is a training‑free online safeguard for generative action‑chunked imitation learning policies. It uses a zero‑shot failure detector based on Temporal Inter‑chunk Discrepancy Estimate (TIDE) and a state‑respawning mechanism that returns the robot to a verified safe intermediate state. The system builds a checkpoint library offline with a vision‑language model and monitors self‑consistency online, rewinding execution to the latest safe checkpoint when a failure is detected, thereby improving reliability in long‑horizon manipulation tasks.

By Gehan Zheng, Sanjay Seenivasan, Matthew Johnson-Roberson, Weiming Zhi

llms robotics multimodal

arXiv AI

Oct 2

Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models

arXiv:2603.16065v3 Announce Type: replace-cross Abstract: Reinforcement Learning (RL) has shown strong potential for improving robotic manipulation policies, yet its practical use remains bottlenecke...

By Yanru Wu, Weiduo Yuan, Esteban Martinez Licon, Ang Qi, Vitor Guizilini, Jiageng Mao, Yue Wang

llms reinforcement-learning robotics fine-tuning multimodal benchmarks

arXiv Machine Learning

Aug 13

Adaptation of Generalist Robot Policies with Minimal Data

arXiv:2608. 11363v1 Announce Type: cross Abstract: A central goal in robot learning is to move beyond task-specific human data collection toward robots that improve through autonomous interaction.

By Shreyas Kowshik, Sreyas Venkataraman, Leo Wang, Niharika Pant, Max Simchowitz, Aviral Kumar

agents reinforcement-learning robotics

arXiv AI

Jun 9

Benchmarking Vision-Language-Action Models on SO-101: Failure and Recovery Analysis

arXiv:2606. 08881v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have demonstrated strong generalization in robotic manipulation, yet existing evaluations are primarily conducted in simulation or on expensive robotic platforms, leaving their robustness on affordable real-world robots largely unexplored.

By Yi Yu, Xinchuan Qiu

robotics fine-tuning multimodal benchmarks

arXiv AI

Sep 18

HIL-UMI: Bringing Human-in-the-Loop Post-Training of Vision-Language-Action Models to Universal Manipulation Interface

HIL-UMI is a policy-guided Universal Manipulation Interface that enables robot‑free, human‑in‑the‑loop post‑training of vision‑language‑action models. By querying the current policy during handheld demonstrations and using an Energy Score to detect out‑of‑distribution states, it selectively collects new data and refines a progress‑based advantage estimator. The updated estimator then drives advantage‑conditioned behavioral cloning, improving performance on long‑horizon and precise manipulation tasks while reducing per‑frame collection time compared to HG‑DAgger.

By Zimu Han, Yiming Zeng, Jiyao Zhang, Zihao Zhao, Yuanfei Wang, Yixiang Jin, Shiqi Li, Shuangben Chen, Wei Huang, Ruodai Li, Hui Shen, Hao Dong

robotics fine-tuning multimodal

One email a morning, machine-written

Related stories

Z-1: Efficient Reinforcement Learning for Vision-Language-Action Models

Rewind-IL: Online Failure Detection and State Respawning for Imitation Learning

Large Reward Models: Generalizable Online Robot Reward Generation with Vision-Language Models

Adaptation of Generalist Robot Policies with Minimal Data

Benchmarking Vision-Language-Action Models on SO-101: Failure and Recovery Analysis

HIL-UMI: Bringing Human-in-the-Loop Post-Training of Vision-Language-Action Models to Universal Manipulation Interface