TissueMicroscopyBench: 49 evals show AI agents lack visual biology intuition under the microscope
anshulkundaje · x · 2026-11-12
kenbwork's team introduced TissueMicroscopyBench, the first release in a new microscopy benchmark series covering tissues, cells and in vitro systems — testing whether agents can do what biologists do at the microscope.
- 49 evaluations built from 4 public imaging datasets.
- Agents must locate structures, trace boundaries, reason over 3D images, and track changes over time.
- Example tasks: outlining mucosa, fat and nerves in H&E-stained colorectal cancer sections, finding evidence of tumor invasion, multiplexed immunofluorescence analysis, liver samples.
- Author chriswzou notes microscopy is the "Hello World" of life sciences, yet agents show poor visual biology intuition.
Takeaway: giving agents microscopy vision is a key but unsolved source of progress in AI for science.
Related event: New Benchmark TissueMicroscopyBench Tests AI's Microscopy Intuition(2 posts)→
More from AGI Musings
- Gary Marcus echoes comparison of Amodei's AI-cures-cancer claim to Theranos — GaryMarcus · 2026-11-12
- Robin Li's three principles for AI-era organizations: builders and sellers — Baidu_Inc · 2026-11-12
- Multi-agent skepticism echoes past agent criticisms that got solved and forgotten — marktenenholtz · 2026-11-12
- Why demand-side aggregators will always block AI agents — and whether agent-native startups are a dumb pipe — signulll · 2026-11-12
- Longevity researcher: those dying in the next decade may be the unluckiest humans ever — rand_longevity · 2026-11-12
- New post argues causal-diagram FDT fails against almost-twins: trivially different agents defect for $5 — jessi_cata · 2026-11-12