Towards Long-Lived Robots: Continual Learning VLA Models via Reinforcement Fine-Tuning

TL;DR

This paper introduces LifeLong-RFT, a reinforcement fine-tuning strategy for vision-language-robotics models that improves continual learning and task adaptation without requiring online feedback or reward models.

Contribution

It proposes a novel reinforcement fine-tuning method with multi-dimensional rewards that enhances multi-task and continual learning for VLA models.

Findings

Achieves a 22% success rate improvement over supervised fine-tuning on LIBERO.

Effectively adapts to new tasks with only 20% of the training data.

Demonstrates strong performance across simulated and real-world tasks.

Abstract

Pretrained on large-scale and diverse datasets, VLA models demonstrate strong generalization and adaptability as general-purpose robotic policies. However, Supervised Fine-Tuning (SFT), which serves as the primary mechanism for adapting VLAs to downstream domains, requires substantial amounts of task-specific data and is prone to catastrophic forgetting. To address these limitations, we propose LifeLong-RFT, a simple yet effective Reinforcement Fine-Tuning (RFT) strategy for VLA models independent of online environmental feedback and pre-trained reward models. By integrating chunking-level on-policy reinforcement learning with the proposed multi-dimensional process reward mechanism, LifeLong-RFT quantifies the heterogeneous contributions of intermediate action chunks across three dimensions to facilitate policy optimization. Specifically, (1) the Quantized Action Consistency Reward (QACR) ensures accurate action prediction within the discrete action space; (2) the Continuous Trajectory Alignment Reward (CTAR) aligns decoded continuous action chunks with reference trajectories to ensure precise control; (3) the Format Compliance Reward (FCR) guarantees the structural validity of outputs. Comprehensive experiments across SimplerEnv, LIBERO, and real-world tasks demonstrate that LifeLong-RFT exhibits strong performance in multi-task learning. Furthermore, for continual learning on the LIBERO benchmark, our method achieves a 22% gain in average success rate over SFT, while effectively adapting to new tasks using only 20% of the training data. Overall, our method provides a promising post-training paradigm for VLAs. The project page is available at <https://yuan-liu-lifelong-rft.github.io>.