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

2026-06-25

AI Summary

Existing test-time scaling methods struggle to jointly scale reasoning and action in embodied tasks and often neglect historical context, limiting long-horizon performance. This work proposes E-TTS, a novel framework that unifies the joint sampling and scoring of reasoning and actions for the first time, incorporating a history buffer mechanism to model temporal dependencies. It further integrates a vision-language verifier with closed-loop feedback to enable environment-adaptive iterative refinement. Designed with modularity, E-TTS flexibly accommodates diverse tasks without requiring additional data or retraining. Extensive experiments across four benchmarks, six environments, three robot platforms, and four foundation models demonstrate performance gains of up to 33.14% in simulation and 26.62% in real-world settings.

Abstract

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.

Problem

• test-time scaling
• embodied tasks
• reasoning
• historical context
• robotic manipulation

Innovation

• Embodied Test-Time Scaling
• History-Aware Reasoning
• Closed-Loop Iterative Refinement
• Vision-Language Verifier
• Modular Robotic Framework