Chain-of-Action: Trajectory Autoregressive Modeling for Robotic Manipulation

TL;DR

This paper introduces Chain-of-Action, a novel visuo-motor policy that generates entire manipulation trajectories through backward reasoning, achieving state-of-the-art results in diverse robotic tasks.

Contribution

The paper proposes a new trajectory autoregressive model with backward reasoning and task-specific goal encoding, enhancing generalization and flexibility in robotic manipulation.

Findings

Achieves state-of-the-art performance on 60 RLBench tasks.

Demonstrates strong generalization in 8 real-world tasks.

Introduces a unified autoregressive structure with backward reasoning.

Abstract

We present Chain-of-Action (CoA), a novel visuo-motor policy paradigm built upon Trajectory Autoregressive Modeling. Unlike conventional approaches that predict next step action(s) forward, CoA generates an entire trajectory by explicit backward reasoning with task-specific goals through an action-level Chain-of-Thought (CoT) process. This process is unified within a single autoregressive structure: (1) the first token corresponds to a stable keyframe action that encodes the task-specific goals; and (2) subsequent action tokens are generated autoregressively, conditioned on the initial keyframe and previously predicted actions. This backward action reasoning enforces a global-to-local structure, allowing each local action to be tightly constrained by the final goal. To further realize the action reasoning structure, CoA incorporates four complementary designs: continuous action token representation; dynamic stopping for variable-length trajectory generation; reverse temporal ensemble; and multi-token prediction to balance action chunk modeling with global structure. As a result, CoA gives strong spatial generalization capabilities while preserving the flexibility and simplicity of a visuo-motor policy. Empirically, we observe CoA achieves the state-of-the-art performance across 60 RLBench tasks and 8 real-world manipulation tasks.