From customer workflows to SpatialBench: applying frontier models to spatial biology
kenbwork · x · 2026-10-10
The author recounts building early prototypes with 2025 reasoning models after studying workflows at spatial, single-cell and proteomics vendors, concluding that code is the substrate for models to manipulate assay data, coding/statistical reasoning transfers broadly across computational biology, and existing coding harnesses suffice given post-training and rigorous evals. This led to SpatialBench, the first spatial biology benchmark, now extended to scRNA, WGS, ATAC, ChIP and concrete drug development activities.
More from AGI Musings
- Pedro Domingos: the worse your memory, the better LLMs seem to you — pmddomingos · 2026-10-10
- OpenAI reportedly solved 92 of the 500 most important open math problems in one GitHub push — altryne · 2026-10-10
- Quanta asks: is AI the end of math as we know it? — burny_tech · 2026-10-10
- AI math is teleportation to a foggy summit: Cepelewicz's striking mountain metaphor — burny_tech · 2026-10-10
- Chollet: science is the ultimate recursively self-improving system, yet progress stays linear — fchollet · 2026-10-10
- Contra Steven Pinker: new article weighs in on Scott Alexander's open-letter feud — repligate · 2026-10-10