Kosmos: An AI Scientist for Autonomous Discovery
2025/11/04 by Ludovico Mitchener, Mitchener, Ludovico, Angela Yiu +75 · 14 voices · 14 citations
Biochemistry, Genetics and Molecular Biology · Decision Sciences · Materials Science · #Biomedical Text Mining and Ontologies #Scientific Computing and Data Management #Machine Learning in Materials Science
paper · pdf · doi:10.48550/arxiv.2511.02824
Abstract
Data-driven scientific discovery requires iterative cycles of literature search, hypothesis generation, and data analysis. Substantial progress has been made towards AI agents that can automate scientific research, but all such agents remain limited in the number of actions they can take before losing coherence, thus limiting the depth of their findings. Here we present Kosmos, an AI scientist that automates data-driven discovery. Given an open-ended objective and a dataset, Kosmos runs for up to 12 hours performing cycles of parallel data analysis, literature search, and hypothesis generation before synthesizing discoveries into scientific reports. Unlike prior systems, Kosmos uses a structured world model to share information between a data analysis agent and a literature search agent. The world model enables Kosmos to coherently pursue the specified objective over 200 agent rollouts, collectively executing an average of 42,000 lines of code and reading 1,500 papers per run. Kosmos cites all statements in its reports with code or primary literature, ensuring its reasoning is traceable. Independent scientists found 79.4% of statements in Kosmos reports to be accurate, and collaborators reported that a single 20-cycle Kosmos run performed the equivalent of 6 months of their own research time on average. Furthermore, collaborators reported that the number of valuable scientific findings generated scales linearly with Kosmos cycles (tested up to 20 cycles). We highlight seven discoveries made by Kosmos that span metabolomics, materials science, neuroscience, and statistical genetics. Three discoveries independently reproduce findings from preprinted or unpublished manuscripts that were not accessed by Kosmos at runtime, while four make novel contributions to the scientific literature.
Citations
Cited by
Discussions
- This is pure bullshit. You don’t need to run 200 LLM agents for >8 hours to do DE, GSEA and generate a couple of heatmaps. And if you do, just spend 2 weeks learning bioinformatics and you won’t need [bsky, 5 points, 2 comments]
- Reading in journal club: Mitchener et al. Kosmos: An AI Scientist for Autonomous Discovery arxiv.org/abs/2511.02824 [bsky, 2 points, 0 comments]
- The preprint seems not-insane: Mitchener et al 2025 doi.org/10.48550/arX... although not in my field. Not 'replace scientists', and not really 'new hypotheses', but I can imagine some big labs might b [bsky, 1 points, 1 comments]
- Kosmos introduces a new era of AI-driven science by autonomously performing literature review, hypothesis generation, and data analysis. It produced reproducible, novel findings across multiple domain [bsky, 1 points, 0 comments]
- Kosmos: An AI Scientist for Autonomous Discovery arxiv.org/abs/2511.02824 [bsky, 0 points, 0 comments]
- Kosmos: An AI Scientist for Autonomous Discovery [bsky, 0 points, 0 comments]
- Kosmos: An AI Scientist for Autonomous Discovery https://arxiv.org/abs/2511.02824 [bsky, 0 points, 0 comments]
- Kosmos: An AI Scientist for Autonomous Discovery https://arxiv.org/abs/2511.02824 [comments] [44 points] [bsky, 0 points, 0 comments]
- ICYMI #Kosmos: An #AI #Scientist for Autonomous Discovery (preprint) (via ArXiv) arxiv.org/abs/2511.02824 [bsky, 0 points, 0 comments]
- Kosmos: An AI Scientist for Autonomous Discovery https://arxiv.org/abs/2511.02824 (https://news.ycombinator.com/item?id=45823358) [bsky, 0 points, 0 comments]
- AI Science Robots Post 10 For details, the Kosmos technical paper is on arXiv: arxiv.org/abs/2511.02824 There’s also an early external evaluation (radiation biology): arxiv.org/abs/2511.13825 [bsky, 0 points, 1 comments]
- arxiv.org/abs/2511.02824 [bsky, 0 points, 0 comments]
- I've decided to devote a bit of life one of the Kosmos AI figures arxiv.org/abs/2511.02824 Let's spend a week on Figure 6 -- where Type 2 Diabetes is linked to SSR1 [bsky, 0 points, 1 comments]
- Kosmos: An AI Scientist for Autonomous Discovery https://arxiv.org/abs/2511.02824 (https://news.ycombinator.com/item?id=45823358) [bsky, 0 points, 0 comments]
Related