arXiv:2606. 25198v2 Announce Type: replace Abstract: Autonomous AI Research promises to accelerate the scientific progress of machine learning.
By Antonis Antoniades, Deepak Nathani, Ritam Saha, Alfonso Amayuelas, Ivan Bercovich, Zhaotian Weng, Vignesh Baskaran, Kunal Bhatia, William Yang Wang
DocuTeam is a mixed‑initiative multi‑agent discussion system that allows both users and agents to start and steer conversations around evolving documents. Agents monitor changes to the document and proactively initiate or redirect discussions, while users can shape the dialogue or adopt agent suggestions. In a within‑subjects study with 20 participants, DocuTeam produced outcomes that were rated as more novel, relevant, and specific compared to a baseline, without increasing cognitive load.
By Heechan Lee, Juhyeon Choi, Tae Soo Kim, Juho Kim, Joseph Seering
arXiv:2608.30047v1 Announce Type: new
Abstract: Recent AI systems promise autonomous scientific discovery, claiming to discover algorithms and produce research papers, yet understanding whether they...
By Shitanshu Bhushan, Yunxiang Zhang, Lu Wang
arXiv:2512. 15011v3 Announce Type: replace Abstract: Artificial intelligence (AI) increasingly generates the very content used to train future AI systems.
By Damian Hodel, Jevin D. West
The paper examines when multi‑agent collaboration is truly beneficial as large language models grow more capable. It finds that multi‑agent systems yield systematic advantages mainly for long‑horizon tasks with sparse dependencies, while single‑agent approaches excel in tightly coupled, sequential workflows. The authors introduce SAIGE, a lightweight, graph‑based collaboration framework that balances context efficiency and performance, demonstrating that adding more agents or deeper recursion does not always improve outcomes.
arXiv:2606. 05178v1 Announce Type: cross Abstract: As AI-driven product development accelerates, the bottleneck is shifting from how we build to what we build.
By Tim Dorn, Saara A. Khan, Julie Mumford