cotomi Act: Learning to Automate Work by Watching You Conference: ACM Conference on AI and Agentic Systems; May 26–29, 2026; San Jose, CA, USA ACM Conference on AI and Agentic Systems (CAIS ’26), May 26–29, 2026, San Jose, CA, USA DOI: 10.1145/3786335.3813203 ISBN: 979-8-4007-2415-2/2026/05 CCS: Computing methodologies Intelligent agents CCS: Human-centered computing Web-based interaction
Masafumi Oyamada, Kunihiro Takeoka, Kosuke Akimoto, Ryoma Obara, Masafumi Enomoto, Haochen Zhang, Daichi Haraguchi, Takuya Tamura · 2026
Read cotomi Act: Learning to Automate Work by Watching You Conference: ACM Conference on AI and Agentic Systems; May 26–29, 2026; San Jose, CA, USA ACM Conference on AI and Agentic Systems (CAIS ’26), May 26–29, 2026, San Jose, CA, USA DOI: 10.1145/3786335.3813203 ISBN: 979-8-4007-2415-2/2026/05 CCS: Computing methodologies Intelligent agents CCS: Human-centered computing Web-based interaction by Masafumi Oyamada, Kunihiro Takeoka, Kosuke Akimoto, Ryoma Obara, Masafumi Enomoto, Haochen Zhang, Daichi Haraguchi, Takuya Tamura (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-03231.txt.
Description
Masafumi Oyamada, Kunihiro Takeoka, Kosuke Akimoto, Ryoma Obara, Masafumi Enomoto, Haochen Zhang, Daichi Haraguchi, and Takuya Tamura present cotomi Act, a browser agent that learns office work by watching its user. The aim is an agent that both runs multi-step web tasks reliably and keeps lasting knowledge of how an organization works.
For execution, lazy observation, compressed history, coarse actions, and best-of-N selection scored above the reported human baseline on a WebArena subset. Separately, passive watching is distilled into task boards and wiki pages in a shared workspace. A controlled test confirmed success rose as that learned knowledge grew.
Catalog record
| CREATOR | Masafumi Oyamada, Kunihiro Takeoka, Kosuke Akimoto, Ryoma Obara, Masafumi Enomoto, Haochen Zhang, Daichi Haraguchi, Takuya Tamura |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03231