mlsanityData to 14 Oct 2026

VLAConf: Calibrated Task-Success Confidence for Vision-Language-Action Models

arXiv 2605.29605 ↗First posted 28 May 2026cs.ROcited by 1

On the map: Vision-language-action models → Failure recovery and robustness

also in 1 more
Abstract (arXiv)

Task-success confidence estimation for Vision-Language-Action (VLA) models provides a crucial task-level signal for monitoring manipulation in open-world environments and supporting downstream decision-making. Existing methods typically construct task-success confidence from action-token probabilities. However, such probabilities are not naturally available in flow-matching policies, limiting their applicability to mainstream flow-matching VLAs. To address this issue, we propose VLAConf, a two-stage representation-level confidence framework that operates on frozen pretrained VLA representations. A step-conditioned Coin-Flip Network learns an uncalibrated inverse success-support score from successful demonstrations, while a low-capacity calibrator fitted on outcome-labeled successful and failed rollouts maps the aggregated score to task-success probability. Experimental results on the LIBERO benchmark demonstrate that VLAConf improves online task-success confidence estimation over alternative approaches. We further demonstrate its utility in selective expert assistance, where confidence-triggered handoffs improve task success over no intervention. Its applicability is also evaluated in real-robot experiments. To access the source code and supplementary videos, visit https://sites.google.com/view/vlaconf.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Failure recovery and robustness, one of 13 directions in Vision-language-action models. Builds on 38 of its 40 references that are in the map. First posted on arXiv 28 May 2026.

Builds on

38 of 40 references.

Closest ancestors · 5

Same direction, newest first

  1. 2026Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment12 Feb
  2. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models15 Oct
  3. 2025LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorization4 Oct
  4. 2025Confidence Calibration in Vision-Language-Action Models23 Jul
  5. 2025SAFE: Multitask Failure Detection for Vision-Language-Action Models11 Jun

Other references · 33

  1. 2017On Calibration of Modern Neural Networkscited by 4,960
  2. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  3. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  4. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  5. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,407
  6. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
  7. 2024Octo: An Open-Source Generalist Robot Policycited by 1,484
  8. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,410
  9. 2018Exploration by Random Network Distillationcited by 993
  10. 2023Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generationcited by 819
  11. 2023Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedbackcited by 812
  12. 2022Teaching Models to Express Their Uncertainty in Wordscited by 751
21 more
  1. 2024INSIDE: LLMs' Internal States Retain the Power of Hallucination Detectioncited by 323
  2. 2026Causal World Modeling for Robot Controlcited by 321
  3. 2023Robots That Ask For Help: Uncertainty Alignment for Large Language Model Plannerscited by 308
  4. 2026Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planningcited by 300
  5. 2024AHA: A Vision-Language-Model for Detecting and Reasoning Over Failures in Robotic Manipulationcited by 125
  6. 2024Consistency Flow Matching: Defining Straight Flows with Velocity Consistencycited by 104
  7. 2024Unpacking Failure Modes of Generative Policies: Runtime Monitoring of Consistency and Progresscited by 69
  8. 2024Real-Time Anomaly Detection and Reactive Planning with Large Language Modelscited by 68
  9. 2024Linguistic Calibration of Long-Form Generationscited by 60
  10. 2025Can We Detect Failures Without Failure Data? Uncertainty-Aware Runtime Failure Detection for Imitation Learning Policiescited by 52
  11. 2025Failure Prediction at Runtime for Generative Robot Policiescited by 42
  12. 2024Multi-Task Interactive Robot Fleet Learning with Visual World Modelscited by 36
  13. 2022A System-Level View on Out-of-Distribution Data in Roboticscited by 33
  14. 2023Flipping Coins to Estimate Pseudocounts for Exploration in Reinforcement Learningcited by 28
  15. 2022Failure Prediction with Statistical Guarantees for Vision-Based Robot Controlcited by 21
  16. 2023Asking for Help: Failure Prediction in Behavioral Cloning through Value Approximationcited by 16
  17. 2020Fighting Failures with FIRE: Failure Identification to Reduce Expert Burden in Intervention-Based Learningcited by 8
  18. 2026RC-NF: Robot-Conditioned Normalizing Flow for Real-Time Anomaly Detection in Robotic Manipulationcited by 8
  19. 2025From Demonstrations to Safe Deployment: Path-Consistent Safety Filtering for Diffusion Policiescited by 5
  20. 2024Improving Predictor Reliability with Selective Recalibrationcited by 3
  21. 2025INSIGHT: INference-time Sequence Introspection for Generating Help Triggers in Vision-Language-Action Modelscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Uncertainty-Gated Exploration Noise Suppresses Task Collapse in Online RL Fine-Tuning of a Flow-Matching Vision-Language-Action Policycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026ReconVLA: An Uncertainty-Guided and Failure-Aware Vision-Language-Action Framework for Robotic Control
  2. 2026Is Success All You Need? Investigating the Impact of Input Perturbations on VLA Behaviour in Tabletop Manipulation Tasks
  3. 2025Evaluating Uncertainty and Quality of Vision-Language-Action-enabled Robots
  4. 2026Benchmarking Vision-Language-Action Models on SO-101: Failure and Recovery Analysis
  5. 2026Uncertainty Quantification for Flow-Based Vision-Language-Action Models
  6. 2026Retrieve-then-Steer: Online Success Memory for Test-Time Adaptation of Generative VLAs
  7. 2026VLA-Corrector: Stage-Aware Observable State Understanding for Prompt-Based Closed-Loop Recovery of Vision-Language-Action Policies
  8. 2026VLA-REPLICA: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models