Intel / article

🔬Causal Models Need Causal Data - Xaira’s X-Cell model for Drug Discovery (Bo Wang & Ci Chu, Chief Discovery Officer & Chief AI Scientist)

Source: latent.space
Author: RJ Honicky
Date: Published Jul 21, 2026

Context

Xaira Therapeutics' X-Cell model is trained to predict how gene activity changes after an intervention. The distinction behind this podcast description is that observing which genes are active together gives different information from deliberately reducing a gene's activity and measuring what follows.

The company's research methods describe training on gene-perturbation experiments across different cellular contexts. X-Cell starts from control-cell populations and predicts gene-expression profiles after a perturbation. Such predictions could help researchers choose interventions to investigate in the laboratory. The output describes cellular responses; establishing a useful treatment still requires evidence beyond that prediction.

Key takeaways · AI-distilled

Xaira broke through its scaling wall by collecting roughly 30x more information per dataset. After that, adding parameters and compute started buying performance again. The fix was running new experiments, not building a bigger model.
CELLxGENE holds 168M cells, each a count over 20-30K genes. It tells you which genes move together, and correlation alone can't say which one drives which, so models trained on it nail cell types and fail at predicting gene edits.
CRISPR lets you turn one gene down at a time and watch what changes downstream. Run millions of those in parallel and you get causal data: A causes B, not just A and B co-vary. That is the X-Atlas dataset underneath X-Cell.
The budget shape is the tell: data collection ran to tens of millions of dollars while compute, headcount and research came to a few million. That inverts the usual ratio and looks more like a reinforcement-learning rollout budget.

Terms in this piece · Glossary

pretraining — The first, biggest phase of building a model: training it on enormous amounts of text so it learns language, facts, and reasoning in general.

Key quotes

“If test loss flatlines after 1.5B parameters while training loss continues to drop as you scale, that tells you that your model is limited by the amount of information in your data.”

“The budget looks like a RL rollout budget, rather than a data rich pre-training one.”

“Models trained on CELLxGENE describe the relationship between cell types and cell states, but they are not good at predicting what will happen if we make changes to RNA expression.”

Read the source www.latent.space