Reproduce “SearchJev: A Calibrated System-1 Model for Search Agents” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Congfeng Cao, Lipeng Zuo, Konstantinos Papakostas, Qiwei Xu, Songwei Xu, Qianru Meng, Lun Zhou, Zhaochun Ren and 2 more

SearchJev separates short, schema-defined search decisions from planning and answer writing: a pretrained Qwen3.5 model scores legal options in one forward pass, while a larger reasoner handles open-ended work and uncertain choices. On 100 BrowseComp-Plus questions, its 4B agent reached 54% judged accuracy versus 45%, with 3.7× faster median active search; results are single-run in this tested setup.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.