Stanford Scholar Questions Efficacy of Single-Cell Perturbation Data for AI

Stanford professor Anshul Kundaje questions the reliance on scPerturb data for scaling single-cell AI models, citing a lack of compelling evidence for its suitability and counterfactual predictive power.

2026-08-06 ~ 2026-08-06 · 2 related posts

• Researcher Questions scPerturb Efficacy as Primary Scaling Axis for Cell Models — anshulkundaje · 2026-08-06
• Expert questions scalability of single-cell data for AI training, citing lack of counterfactual predictive power — anshulkundaje · 2026-08-06