![[2602.21201] Aletheia tackles FirstProof autonomously](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.21201] Aletheia tackles FirstProof autonomously
Summary
The paper presents Aletheia, an autonomous mathematics research agent that successfully solved 6 out of 10 problems in the FirstProof challenge, showcasing advancements in AI's problem-solving capabilities.
Why It Matters
This research highlights the potential of AI agents in tackling complex mathematical problems autonomously, marking a significant step forward in AI applications in mathematics and potentially influencing future research methodologies and educational tools.
Key Takeaways
- Aletheia autonomously solved 6 out of 10 problems in the FirstProof challenge.
- Expert assessments varied, particularly on Problem 8, indicating areas for improvement.
- The study emphasizes transparency in AI evaluations and problem interpretations.
Computer Science > Artificial Intelligence arXiv:2602.21201 (cs) [Submitted on 24 Feb 2026] Title:Aletheia tackles FirstProof autonomously Authors:Tony Feng, Junehyuk Jung, Sang-hyun Kim, Carlo Pagano, Sergei Gukov, Chiang-Chiang Tsai, David Woodruff, Adel Javanmard, Aryan Mokhtari, Dawsen Hwang, Yuri Chervonyi, Jonathan N. Lee, Garrett Bingham, Trieu H. Trinh, Vahab Mirrokni, Quoc V. Le, Thang Luong View a PDF of the paper titled Aletheia tackles FirstProof autonomously, by Tony Feng and 16 other authors View PDF HTML (experimental) Abstract:We report the performance of Aletheia (Feng et al., 2026b), a mathematics research agent powered by Gemini 3 Deep Think, on the inaugural FirstProof challenge. Within the allowed timeframe of the challenge, Aletheia autonomously solved 6 problems (2, 5, 7, 8, 9, 10) out of 10 according to majority expert assessments; we note that experts were not unanimous on Problem 8 (only). For full transparency, we explain our interpretation of FirstProof and disclose details about our experiments as well as our evaluation. Raw prompts and outputs are available at this https URL. Comments: Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) Cite as: arXiv:2602.21201 [cs.AI] (or arXiv:2602.21201v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2602.21201 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Thang Luong [view email] [v1] Tue, ...