الملخص
GUI agents are deployed with frozen weights and discard everything they experience on the job. Existing ways to update an agent's weights assume something deployment withholds: ground truth, rollouts beyond the single attempt (retries, samples, practice runs), or a learning phase other than deployment. Because GUI actions can be irreversible, a deployed agent gets one attempt per task occurrence, in arrival order, and every attempt counts. No ground truth is available at any point. We define fully test-time adaptation for GUI agents by these constraints and pair it with a minimal weight-space method, SOLO. Auxiliary models read each episode: a judge selects the episodes it deems successful, and a proposer-verifier pair relabels a failed episode's prefix with the subtask that prefix completed. Admitted episodes enter a short sliding window, and each admission updates a small adapter by top-K self-distillation on the agent's own predictions, provided the window holds a judged success. On recurring task streams built from WebArena, VisualWebArena and MobileWorld, SOLO 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.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Wang, Z., Gu, L., Chi, Z., Jiang, L., Jiang, Z., Guo, L., Reid, S., Liu, Z., & Wang, Y. (2026). One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents. https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents
MLA 9
Wang, Ziqiang, et al. "One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents." https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents.
شيكاغو (المؤلف–التاريخ)
Wang, Ziqiang, Li Gu, Zhixiang Chi, Linlian Jiang, Zihuan Jiang, Linqiang Guo, Siobhan Reid, Zhi Liu, and Yang Wang. 2026. "One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents." https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents.
هارفارد
Wang, Z., Gu, L., Chi, Z., Jiang, L., Jiang, Z., Guo, L., Reid, S., Liu, Z. and Wang, Y. (2026) 'One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents', Available at: https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents.
فانكوفر
Wang Z, Gu L, Chi Z, Jiang L, Jiang Z, Guo L, et al. One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents. https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents
IEEE
Z. Wang, L. Gu, Z. Chi, L. Jiang, Z. Jiang, L. Guo, S. Reid, Z. Liu, and Y. Wang, "One Rollout Is All You Get: Fully Test-Time Adaptation for GUI Agents," https://omanscience.com/ar/articles/one-rollout-is-all-you-get-fully-test-time-adaptation-for-gui-agents.