Kundaje: best-performing perturb models (PRESAGE, PIE, Rheister) use no scFM pretraining
anshulkundaje · x · 2026-10-06
In post 7 of his thread, Anshul Kundaje lists the perturbation prediction models that actually do somewhat OK — PRESAGE, Rheister, Arc's new PIE model, and GenbioAI's nearest-neighbor models — and points out that none of them use any single-cell foundation model pretraining on observational atlases. The accompanying post 8 adds that most of the SOTA models' predictive power comes from prior knowledge, so they are not causal transcription regulation models, virtual cells, or world models.
More from Research
- User Sim Index is broken: trivial bot scores 95% across behavioral dims — ericzelikman · 2026-10-06
- Used OpenAI Dots as a Free Agent Swarm to Break a 47-Year-Old Math Record — jaxchang · 2026-10-06
- Stanford prof says current experiment designs can't train causal virtual cell models — anshulkundaje · 2026-10-06
- TIDES dataset on multi-party and multi-agent collaboration to debut at COLM 2026 — josephseering · 2026-10-06
- LakeQuest QA benchmark, testing RAG on messy enterprise tables and docs, hits COLM — hllo_wrld · 2026-10-06
- Frontier Data Summit lineup: Chollet, Dawn Song headline batch of new agent benchmarks — StanfordAILab · 2026-10-06