An AI agent for treatment reasoning over a biomedical tool universe
Quick Answer
ATHENA-R1 is an AI agent for treatment reasoning, outperforming existing models with 94.7% accuracy in drug reasoning and 82.9% in treatment reasoning.
Quick Take
Trained using reinforcement learning across 3,168 drug tasks and 456 patient cases, it shows significant improvements over GPT-5 by 17.8 and 10.7 points respectively.
Key Points
- ATHENA-R1 integrates 212 biomedical tools for comprehensive treatment reasoning.
- Achieved 94.7% accuracy on drug reasoning tasks across five benchmarks.
- Outperformed GPT-5 by 17.8 points in drug reasoning accuracy.
- Preferred by experts over reference models in evaluations from 28 rare disease organizations.
- Generated adverse-event hypotheses with adjusted odds ratios of 1.48-1.84.
Paper Resources
📖 Reader Mode
~2 min readAuthors:Shanghua Gao, Ayush Noori, Richard Zhu, Curtis Ginder, Zhenglun Kong, Xiaorui Su, Justin Kauffman, Benjamin S. Glicksberg, Joshua Lampert, Ankit Sakhuja, Ashwin Sawant, ATHENA-R1 Evaluation Consortium, David A. Clifton, Noa Dagan, Ran Balicer, Marinka Zitnik
Abstract:Treatment reasoning underpins every therapeutic decision, integrating disease context, comorbidities, medications, contraindications, and evolving biomedical knowledge to select an appropriate therapy. It is inherently iterative: candidates are weighed against many constraints, revised as evidence emerges, and grounded in verifiable sources. Here we introduce ATHENA-R1, an AI agent for treatment reasoning across all FDA approved drugs since 1939, trained by reinforcement learning over a universe of 212 biomedical tools. At each step it identifies missing information, selects and runs relevant tools, and incorporates the evidence. To train it without human-annotated traces, we build a two-level self-learning framework: multi-agent systems construct the tools, tasks, and reasoning trajectories for supervised fine-tuning, then reinforcement learning with scientific feedback rewards reasoning quality (evidence gathering, grounded tool use, logical non-redundancy). Across five benchmarks of 3,168 drug reasoning tasks and 456 patient treatment cases, ATHENA-R1 outperforms language models and tool-use systems, reaching 94.7% accuracy on open-ended drug reasoning and 82.9% on treatment reasoning, 17.8 and 10.7 points above GPT-5. In blinded evaluations by experts from 28 rare disease organizations, it is preferred over reference models on all criteria, and physicians rated it favorably on complex hospitalized cardiovascular and infectious-disease cases. Adverse-event hypotheses it generated, tested in electronic health records from 5.4 million patients, reached adjusted odds ratios of 1.48-1.84, with no elevation among negative controls. Because it requires knowing what evidence to seek before concluding, treatment reasoning has long been hard for AI; we show it can be reframed as a learnable process of iterative evidence gathering that reinforcement learning can train AI to perform.
| Comments: | Project page: this https URL Code: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.28692 [cs.AI] |
| (or arXiv:2606.28692v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2606.28692 arXiv-issued DOI via DataCite |
Submission history
From: Shanghua Gao [view email]
[v1]
Sat, 27 Jun 2026 02:24:56 UTC (8,433 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.