# Microsoft — Beyond the benchmark: How an adaptive approach drives scientific discovery

- Company: Microsoft (microsoft.com)
- Announced: 2026-09-08T20:00:00+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://azure.microsoft.com/en-us/blog/beyond-the-benchmark-how-an-adaptive-approach-drives-scientific-discovery/
- Record: https://forck.live/items/10463-beyond-the-benchmark-how-an-adaptive-approach-drives-scientific-discovery
- Subject: Azure AI

Microsoft announced that its Discovery Engine with CLIO achieved scores of 61.6% in health and medicine, 75.2% in physical sciences, and 64.6% in life sciences on the Agent's Last Exam benchmark. The system uses adaptive reasoning to explore multiple hypotheses and refine approaches. Microsoft Discovery is available as an enterprise platform for agentic R&D.

## Evidence

Verbatim from https://azure.microsoft.com/en-us/blog/beyond-the-benchmark-how-an-adaptive-approach-drives-scientific-discovery/:

> On Agent’s Last Exam, a demanding evaluation of long-running, tool-using professional tasks, Microsoft Discovery Engine with CLIO (Cognitive Loop via In-Situ Optimization) achieved higher scores than the other agentic harnesses evaluated across three scientific domains: 61.6% in health and medicine, 75.2% in physical sciences, and 64.6% in life sciences.

---

Record: https://forck.live/items/10463-beyond-the-benchmark-how-an-adaptive-approach-drives-scientific-discovery
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md