Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.18197 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914280423555072 |
|---|---|
| author | Wang, Shaokang Fu, Pei Zhang, Ruoceng Zhang, Shaojie Xi, Xiuwen Yang, Jiahui Qin, Bin Huang, Ying Luo, Zhenbo Luan, Jian |
| author_facet | Wang, Shaokang Fu, Pei Zhang, Ruoceng Zhang, Shaojie Xi, Xiuwen Yang, Jiahui Qin, Bin Huang, Ying Luo, Zhenbo Luan, Jian |
| contents | While Large Vision-Language Models (LVLMs) have significantly advanced GUI agents' capabilities in parsing textual instructions, interpreting screen content, and executing tasks, a critical challenge persists: the irreversibility of agent operations, where a single erroneous action can trigger catastrophic deviations. To address this, we propose the GUI Action Critic's Data Flywheel System (GAIA), a training framework that enables the models to have iterative critic capabilities, which are used to improve the Test-Time Scaling (TTS) of basic GUI agents' performance. Specifically, we train an Intuitive Critic Model (ICM) using positive and negative action examples from a base agent first. This critic evaluates the immediate correctness of the agent's intended actions, thereby selecting operations with higher success probability. Then, the initial critic guides agent actions to collect refined positive/negative samples, initiating the self-improving cycle. The augmented data then trains a second-round critic with enhanced discernment capability. We conduct experiments on various datasets and demonstrate that the proposed ICM can improve the test-time performance of various closed-source and open-source models, and the performance can be gradually improved as the data is recycled. The code and dataset will be publicly released. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18197 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | GAIA: A Data Flywheel System for Training GUI Test-Time Scaling Critic Models Wang, Shaokang Fu, Pei Zhang, Ruoceng Zhang, Shaojie Xi, Xiuwen Yang, Jiahui Qin, Bin Huang, Ying Luo, Zhenbo Luan, Jian Artificial Intelligence While Large Vision-Language Models (LVLMs) have significantly advanced GUI agents' capabilities in parsing textual instructions, interpreting screen content, and executing tasks, a critical challenge persists: the irreversibility of agent operations, where a single erroneous action can trigger catastrophic deviations. To address this, we propose the GUI Action Critic's Data Flywheel System (GAIA), a training framework that enables the models to have iterative critic capabilities, which are used to improve the Test-Time Scaling (TTS) of basic GUI agents' performance. Specifically, we train an Intuitive Critic Model (ICM) using positive and negative action examples from a base agent first. This critic evaluates the immediate correctness of the agent's intended actions, thereby selecting operations with higher success probability. Then, the initial critic guides agent actions to collect refined positive/negative samples, initiating the self-improving cycle. The augmented data then trains a second-round critic with enhanced discernment capability. We conduct experiments on various datasets and demonstrate that the proposed ICM can improve the test-time performance of various closed-source and open-source models, and the performance can be gradually improved as the data is recycled. The code and dataset will be publicly released. |
| title | GAIA: A Data Flywheel System for Training GUI Test-Time Scaling Critic Models |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2601.18197 |