mlsanityData to 14 Oct 2026

ACRO: Actor-Critic Rollout Orchestration

arXiv 2610.19677 ↗First posted 14 Oct 2026cs.ROcited by 0

On the map: Vision-language-action models → Reinforcement learning and post-training also fits: Failure recovery and robustness

also in 1 more
Abstract (arXiv)

Executing a vision-language-action (VLA) policy involves deciding when to continue an attempt and when to intervene to improve its chances of success. We introduce ACRO, an Actor--Critic rollout orchestration layer that makes these decisions for a fixed VLA. ACRO seeks interventions that offer a more promising continuation, are physically feasible, and leave sufficient time to complete the task. A Critic learned from policy rollouts estimates success prospects to decide when to intervene and select a re-entry target along the executed trajectory. Trajectory Retraction shortcuts the visited path to return the robot to that target, where the same VLA resumes execution from a fresh observation. We evaluate ACRO with GR00T N1.7 and π0.5 across SimplerEnv, RoboCasa, and real-robot manipulation. Under the same physical execution budget, including all return motion, ACRO improves success by 10.4, 7.3, and 21.7 percentage points, respectively. It also completes additional tasks as the execution budget grows. These results show that controlling when and where a policy resumes can turn its existing capabilities into more successful task executions. Project page: https://junhoo.me/acro

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Reinforcement learning and post-training, one of 13 directions in Vision-language-action models. Builds on 13 of its 19 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

13 of 19 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Steering Your Generalists: Improving Robotic Foundation Models via Value Guidance17 Oct

Other references · 12

  1. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  2. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  3. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,407
  4. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,348
  5. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,761
  6. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 495
  7. 2025SAFE: Multitask Failure Detection for Vision-Language-Action Modelscited by 80
  8. 2019Continuous and Discrete-Time Survival Prediction with Neural Networkscited by 70
  9. 2025Failure Prediction at Runtime for Generative Robot Policiescited by 42
  10. 2026Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignmentcited by 21
  11. 2026CycleVLA: Proactive Self-Correcting Vision-Language-Action Models via Subtask Backtracking and Minimum Bayes Risk Decodingcited by 14
  12. 2026Addressing the Orchestration Gap in Generalist Robots via Physical Agencycited by 11

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026PAPO-VLA: Planning-Aware Policy Optimization for Vision-Language-Action Models
  2. 2026DriftOPD: Sequence-Level Reverse-KL Distillation for One-Step VLA Policies
  3. 2026Prism-GRPO: Faster VLA Policy Optimization via Splitting Same-outcome Groups
  4. 2026SWAP: Stepwise Action Policy Routing for Vision-Language-Action Models
  5. 2026Find Something You Can't Do: Agentic Real-World Reinforcement Learning for Self-Improving VLA Models
  6. 2026RePO-VLA: Recovery-Driven Policy Optimization for Vision-Language-Action Models
  7. 2026VLA Knows Its Limits: Adaptive Execution Horizons for Robot Policies
  8. 2026Fewer Steps, Better Actions: Rethinking Flow-Matching Inference for VLA Policies