arXiv cs.AI· Pierre-Louis Favreau, Jean-Pierre Lo, Clement Guiguet, Charles Simon-Meunier, Nicolas Dehandschoewercker, Allen G. Roush, Judah Goldfeder, Ravid Shwartz-Ziv·· 4 小时前AI 评分58
Minitap 多智能体系统在 AndroidWorld 上通过任务分解达到 100% 通过率
Do Multi-Agents Dream of Electric Screens? Achieving 100% on AndroidWorld Through Task Decomposition
AI 导读
Minitap 团队提出多智能体系统 Minitap's mobile-use,在 AndroidWorld 的 116 个任务上取得 pass@3 100.0%(95% CI 96.8-100.0),pass^3 为 94.0%,平均成功率 98.0%。
正文 · 原文
Abstract:We present Minitap's mobile-use, a multi-agent system for Android UI control, and evaluate it on the 116 tasks of AndroidWorld, on Android emulators in Minitap's cloud device infrastructure, real Pixel 7 phones from our device farm and the stock emulator. On fresh, held-out task instances (three pre-registered seeds never used during development), the final system reaches a pass@3 of 100.0% (95% CI 96.8-100.0); pass^3 is 94.0% (95% CI 89.7-97.4) and the mean success rate 98.0% (95% CI 96.0-99.7). Minitap decomposes each task across five LLM agents (planning, subgoal tracking, state capture, decision, execution) and deterministic nodes, with mechanisms that read typed text back from the device, verify the goal on device evidence and detect repeated failures. A study of the initial system measures what each part contributes: removing only the decomposition changes success by -26.1 [-35.6, -16.7] points, the mechanisms together add +11.7 [+2.2, +21.1] points over a single agent, and the decision agent is the one role that needs a frontier model. We then improve reliability with general changes, a few of them triggered by the goal's wording. On MobileMiniWoB++, a second benchmark never used during development, the final system improves on the initial one by 7.1 points (McNemar p = 0.035; 95% CI -0.5 to +14.7). Minitap is open source: this https URL
| Comments: | Accepted at NeurIPS 2026 Agents in the Wild: Safety, Security, and Beyond |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2602.07787 [cs.AI] |
| (or arXiv:2602.07787v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2602.07787 arXiv-issued DOI via DataCite |
Submission history
From: Judah Goldfeder [view email]
[v1]
Sun, 8 Feb 2026 03:02:13 UTC (5,935 KB)
[v2]
Mon, 16 Nov 2026 21:53:41 UTC (1,002 KB)
来源:arXiv cs.AI · arxiv.org