SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models

Hyeonbeom Choi, Daechul Ahn, Youhan Lee, Taewook Kang, Seongwon Cho, Jonghyun Choi
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:19922-19943, 2026.

Abstract

Vision-Language-Action (VLA) models have emerged as a promising paradigm for general-purpose robotic control, with test-time scaling (TTS) gaining attention to enhance robustness beyond training. However, existing TTS methods for VLAs require additional training, verifiers, and multiple forward passes, making them impractical for deployment. Moreover, they intervene only at action decoding while keeping visual representations fixed—insufficient under perceptual ambiguity, where reconsidering how to perceive is as important as deciding what to do. To address these limitations, we propose SCALE, a simple inference strategy that jointly modulates visual perception and action based on ’self-uncertainty’, inspired by uncertainty-driven exploration in Active Inference theory—requiring no additional training, no verifier, and only a single forward pass. SCALE broadens exploration in both perception and action under high uncertainty, while focusing on exploitation when confident—enabling adaptive execution across varying conditions. Experiments on simulated and real-world benchmarks demonstrate that SCALE improves state-of-the-art VLAs and outperforms existing TTS methods while maintaining single-pass efficiency. Our code is publicly available at https://github.com/snumprlab/scale.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-choi26a, title = {{SCALE}: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models}, author = {Choi, Hyeonbeom and Ahn, Daechul and Lee, Youhan and Kang, Taewook and Cho, Seongwon and Choi, Jonghyun}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {19922--19943}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/choi26a/choi26a.pdf}, url = {https://proceedings.mlr.press/v306/choi26a.html}, abstract = {Vision-Language-Action (VLA) models have emerged as a promising paradigm for general-purpose robotic control, with test-time scaling (TTS) gaining attention to enhance robustness beyond training. However, existing TTS methods for VLAs require additional training, verifiers, and multiple forward passes, making them impractical for deployment. Moreover, they intervene only at action decoding while keeping visual representations fixed—insufficient under perceptual ambiguity, where reconsidering how to perceive is as important as deciding what to do. To address these limitations, we propose SCALE, a simple inference strategy that jointly modulates visual perception and action based on ’self-uncertainty’, inspired by uncertainty-driven exploration in Active Inference theory—requiring no additional training, no verifier, and only a single forward pass. SCALE broadens exploration in both perception and action under high uncertainty, while focusing on exploitation when confident—enabling adaptive execution across varying conditions. Experiments on simulated and real-world benchmarks demonstrate that SCALE improves state-of-the-art VLAs and outperforms existing TTS methods while maintaining single-pass efficiency. Our code is publicly available at https://github.com/snumprlab/scale.} }
Endnote
%0 Conference Paper %T SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models %A Hyeonbeom Choi %A Daechul Ahn %A Youhan Lee %A Taewook Kang %A Seongwon Cho %A Jonghyun Choi %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-choi26a %I PMLR %P 19922--19943 %U https://proceedings.mlr.press/v306/choi26a.html %V 306 %X Vision-Language-Action (VLA) models have emerged as a promising paradigm for general-purpose robotic control, with test-time scaling (TTS) gaining attention to enhance robustness beyond training. However, existing TTS methods for VLAs require additional training, verifiers, and multiple forward passes, making them impractical for deployment. Moreover, they intervene only at action decoding while keeping visual representations fixed—insufficient under perceptual ambiguity, where reconsidering how to perceive is as important as deciding what to do. To address these limitations, we propose SCALE, a simple inference strategy that jointly modulates visual perception and action based on ’self-uncertainty’, inspired by uncertainty-driven exploration in Active Inference theory—requiring no additional training, no verifier, and only a single forward pass. SCALE broadens exploration in both perception and action under high uncertainty, while focusing on exploitation when confident—enabling adaptive execution across varying conditions. Experiments on simulated and real-world benchmarks demonstrate that SCALE improves state-of-the-art VLAs and outperforms existing TTS methods while maintaining single-pass efficiency. Our code is publicly available at https://github.com/snumprlab/scale.
APA
Choi, H., Ahn, D., Lee, Y., Kang, T., Cho, S. & Choi, J.. (2026). SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:19922-19943 Available from https://proceedings.mlr.press/v306/choi26a.html.

Related Material