Guardado en:
Detalles Bibliográficos
Autores principales: Oyamada, Masafumi, Takeoka, Kunihiro, Akimoto, Kosuke, Obara, Ryoma, Enomoto, Masafumi, Zhang, Haochen, Haraguchi, Daichi, Tamura, Takuya
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2605.03231
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913088972783616
author Oyamada, Masafumi
Takeoka, Kunihiro
Akimoto, Kosuke
Obara, Ryoma
Enomoto, Masafumi
Zhang, Haochen
Haraguchi, Daichi
Tamura, Takuya
author_facet Oyamada, Masafumi
Takeoka, Kunihiro
Akimoto, Kosuke
Obara, Ryoma
Enomoto, Masafumi
Zhang, Haochen
Haraguchi, Daichi
Tamura, Takuya
contents What if a browser agent could learn your work simply by watching you do it? We present cotomi Act, a browser-based computer-using agent that combines reliable multi-step task execution with persistent organizational knowledge learned from user behavior. For execution, an agent scaffold with adaptive lazy observation, verbal-diff-based history compression, coarse-grained actions, and test-time scaling via best-of-N action selection achieves 80.4% on the 179-task WebArena human-evaluation subset, exceeding the reported 78.2% human baseline. For organizational knowledge, a behavior-to-knowledge pipeline passively observes the user's browsing and progressively abstracts it into artifacts (task boards, wiki) exposed through a shared workspace editable by both user and agent. A controlled proxy evaluation confirms that task success improves as behavior-derived knowledge accumulates. In our live demonstration, attendees interact with the system in a real browser, issuing tasks and observing end-to-end autonomous execution and shared knowledge management.
format Preprint
id arxiv_https___arxiv_org_abs_2605_03231
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle cotomi Act: Learning to Automate Work by Watching You
Oyamada, Masafumi
Takeoka, Kunihiro
Akimoto, Kosuke
Obara, Ryoma
Enomoto, Masafumi
Zhang, Haochen
Haraguchi, Daichi
Tamura, Takuya
Artificial Intelligence
What if a browser agent could learn your work simply by watching you do it? We present cotomi Act, a browser-based computer-using agent that combines reliable multi-step task execution with persistent organizational knowledge learned from user behavior. For execution, an agent scaffold with adaptive lazy observation, verbal-diff-based history compression, coarse-grained actions, and test-time scaling via best-of-N action selection achieves 80.4% on the 179-task WebArena human-evaluation subset, exceeding the reported 78.2% human baseline. For organizational knowledge, a behavior-to-knowledge pipeline passively observes the user's browsing and progressively abstracts it into artifacts (task boards, wiki) exposed through a shared workspace editable by both user and agent. A controlled proxy evaluation confirms that task success improves as behavior-derived knowledge accumulates. In our live demonstration, attendees interact with the system in a real browser, issuing tasks and observing end-to-end autonomous execution and shared knowledge management.
title cotomi Act: Learning to Automate Work by Watching You
topic Artificial Intelligence
url https://arxiv.org/abs/2605.03231