Revisiting Embodied Chain-of-Thought for Generalizable Robot Manipulation
cs.RO
Submitted: 2026-06-02
Updated: 2026-10-15
Project page: https://taoshuaiz.github.io/ERVLA
License: http://creativecommons.org/licenses/by/4.0/
The gist: Embodied chain-of-thought (CoT) aims to bridge linguistic reasoning with robotic control, yet its effective form and integration remain underexplored.
Terminology
Abstract
Embodied chain-of-thought (CoT) aims to bridge linguistic reasoning with robotic control, yet its effective form and integration remain underexplored. In this paper, we revisit embodied CoT for robotic control at an unprecedented scale. We curate the largest embodied CoT corpus to date, comprising 978,743 trajectories, 226.3M samples, and 2592.5 hours of data. Through extensive experiments, we show that effective CoT must ground high-level semantic understaning in concrete linguistic action guidance -- such as end-effector movement and image-space trajectories -- whereas high-level reasoning alone yields marginal gains. More importantly, we identify that explicit CoT does not scale reliably as an autoregressive action prefix, suffering from compounding errors during inference. To address these challenges, we propose ERVLA, a vision-language-action (VLA) model that effectively leverages linguistic reasoning in generalizable robot manipulation. ERVLA is trained using a CoT-dropout strategy, allowing the model to leverage rich reasoning traces during training while predicting actions directly without CoT during inference to bypass autoregressive instability. This approach enables reliable scaling with increasing pre-training data. ERVLA achieves state-of-the-art results on LIBERO-Plus with an 86.9% success rate and reaches 53.2% on VLABench, showcasing superior performance in out-of-distribution settings. Furthermore, ERVLA outperforms competitive state-of-the-art baselines in real-robot experiments, especially in handling semantic ambiguity and long-horizon tasks. Code, data, and model checkpoints are available on the project website: https://taoshuaiz.github.io/ERVLA/.
Sources
- Qwen3-VL Technical Report
- PaliGemma: A versatile 3B VLM for transfer
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
- Real-Time Execution of Action Chunking Flow Policies
- UniVLA: Learning to Act Anywhere with Task-centric Latent Actions
- Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution
- WorldVLA: Towards Autoregressive Action World Model
- Training Strategies for Efficient Embodied Reasoning
- StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing
- Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better
- LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models
- LLMDet: Learning Strong Open-Vocabulary Object Detectors under the Supervision of Large Language Models
- ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning
- Fast-ThinkAct: Efficient Vision-Language-Action Reasoning via Verbalizable Latent Planning
- $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
- DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
- Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics
- OpenVLA: An Open-Source Vision-Language-Action Model
- Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
- Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies