Penadés et al. 2025 give the hypothesis-generation claims of Gottweis et al. 2026's Co-Scientist a contamination-controlled test: they posed a question that took years to resolve experimentally but remained unpublished, and the system's top-ranked hypothesis matched the confirmed mechanism—cf-PICIs hijacking diverse phage tails—so the answer cannot have been retrieved from the published literature.
E36 states the question took years to resolve experimentally but remained unpublished and that the AI co-scientist's top-ranked hypothesis matched the confirmed mechanism (cf-PICIs hijacking diverse phage tails), directly implying the answer could not be retrieved from published literature, and E5 identifies Co-Scientist as the multi-agent hypothesis-generation system being tested.
Written by Kimi K3 via Ollama Cloud · checked by GLM-5.3 via Ollama Cloud · 15 Oct, 04:34
Source chain
AEvery quote below was checked, without a model, to appear verbatim in its source.
- 01
“the AI co-scientist's top-ranked hypothesis matched our experimentally confirmed mechanism: cf-PICIs hijack diverse phage tails to expand their host range”
Abstract · no page number · evidence E36
- 02
“it formulates demonstrably novel research hypotheses for experimental verification”
Full-text passage · no page number · evidence E5
Accelerating scientific discovery with Co-Scientist Scientific discovery is driven by scientists generating hypotheses for complex problems that undergo rigorous experimental validation. To augment this process, we introduce Co-Scientist, a multi-agent artificial intelligence (AI) system built on Gemini for structured scientific thinking and hypothesis generation. Co-Scientist aims to help scientists discover new original knowledge. Conditioned on their research objectives and previous scientific evidence, it formulates demonstrably novel research hypotheses for experimental verification. The system’s design involves agents continuously generating, critiquing and refining hypotheses accelerated by scaling test-time compute. Key contributions include (1) a multi-agent architecture with an asynchronous task execution framework for flexible compute scaling, and (2) a tournament evolution process for self-improving hypotheses generation. Automated evaluations show continued benefits of test-time compute scaling, improving hypothesis quality over time.
Passage read from www.ebi.ac.uk, which may be a preprint rather than the published version.