E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulation

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Wen Ye, Peiyan Li, Tingyu Yuan, Yuan Xu, Xiangnan Wu, Chaoyang Zhao, Jing Liu, Nianfeng Liu, Yan Huang, Liang Wang
• Venue: arXiv: cs.RO
• Published: 2026-06-25
• Collected: 2026-06-26
• URL: http://arxiv.org/abs/2606.27268v1
• PDF: https://arxiv.org/pdf/2606.27268v1
• DOI: UNKNOWN

Plain-Language Brief

The researchers introduced a new framework called E-TTS for robotic manipulation that combines reasoning and action scaling using historical context. Unlike traditional open-loop TTS methods, E-TTS includes feedback generation to improve efficiency and adaptability. The authors tested the framework across various benchmarks, environments, embodiments, and models and found consistent performance improvements with up to a 33.14% increase in simulation and 26.62% in real-world scenarios.

What the Document Says

• E-TTS is a new framework for robotic manipulation that combines reasoning and action scaling using historical context.
• It utilizes a history buffer to store historical context, which is used by reasoning and action verifiers to evaluate sampled candidates.
• Unlike traditional open-loop TTS methods, E-TTS includes feedback generation to form a closed-loop iterative refinement mechanism.
• The framework was tested across 4 different benchmarks, 6 environments, 3 embodiments, and 4 base vision-language-action models.
• Experimental results show consistent performance improvements with up to a 33.14% increase in simulation and 26.62% in real-world scenarios.

Key Concepts Mentioned

• E-TTS framework
• Robotic manipulation
• Reasoning and action scaling
• Historical context
• Feedback generation
• Closed-loop iterative refinement

Claims Made by the Document

• E-TTS improves the performance of robotic manipulation.
• It achieves up to a 33.14% increase in simulation and 26.62% in real-world scenarios.

Evidence Used

• The experimental results presented in the text

Methods / Approach

• Reasoning-action joint sampling and scoring
• History buffer for storing historical context
• Feedback generation for closed-loop iterative refinement

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

The article does not specify any limitations of the authors' work.

Open Questions

• What specific techniques were used for reasoning and action scaling in E-TTS?

Notes for Future Q&A

May include interpretation. Labeled where it does.

The text focuses on presenting the framework and its results but does not provide detailed explanations of the methods used.

Raw Abstract / Excerpt

Recently, a few works have made early attempts to study test-time scaling for embodied tasks. However, two major challenges remain unsolved: (1) reasoning can effectively improve the performance of the policy, but its scaling mechanism has seldom been studied; (2) historical information is essential, as embodied tasks are inherently long-horizon and sequential, making sole reliance on current observations for action scaling inadequate due to the lack of historical context utilization. To address these challenges, we introduce E-TTS, a modular and plug-and-play Embodied Test-Time Scaling framework that unifies reasoning and action scaling for robotic manipulation via history-aware iterative refinement with vision-language verifiers. To support joint reasoning-action scaling, E-TTS performs reasoning-action joint sampling and scoring in a pairwise manner. To better utilize historical information, E-TTS uses a history buffer to store historical context, which is then used by reasoning and action verifiers to evaluate the sampled candidates. Unlike conventional open-loop TTS methods, E-TTS introduces feedback generation into the sampling process to form a closed-loop iterative refinement mechanism, enhancing both inference efficiency and environmental adaptability. Each component functions as an independent and composable module, allowing flexible and adaptive configuration depending on task requirements. To evaluate the advantages of our framework, we conduct experiments across 4 different benchmarks, 6 environments, 3 embodiments, and 4 base vision-language-action models. The experimental results demonstrate that, without requiring additional expert data collection or retraining, E-TTS consistently improves performance, achieving up to a 33.14% increase in simulation and 26.62% in real-world scenarios.