This work defines fully test-time adaptation for GUI agents by these constraints and pair it with a minimal weight-space method, SOLO, which improves on the frozen agent with both UI-TARS-7B and Qwen3-VL-8B, by three to six points of success rate, and exceeds two in-setting memory methods on the web streams.
Zi-Qiang Wang, Li Gu, Zhixiang Chi et al.· arXiv.org· 0 citations
This work introduces MobileJudgeBench, a benchmark for systematically evaluating LLM-as-judge methods on mobile agent trajectories, and reveals benchmark quality metrics reliably predict real-world judge utility.
Zi-Qiang Wang, Li Gu, Zhixiang Chi et al.· 0 citations
A new FSTT-DA framework that integrates LoRA fine-tuning with model merging and proposes a hypernetwork trained via meta-learning that generates per-column merging factors to combine LoRA modules to adapt the learned knowledge to a specific target domain.
Siobhan Reid, Zhixiang Chi, Li Gu et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.