New CORE method boosts LLM accuracy in cellular perturbation prediction

A new research paper introduces CORE, a method designed to improve the accuracy of large language models (LLMs) in predicting cellular responses to perturbations. The study found that existing LLM approaches, while appearing plausible, often fail to accurately predict gene expression changes and can underperform simpler baselines. CORE addresses this by reframing the prediction task as a comparative one, organizing evidence from related perturbations to highlight differences in their effects on genes. This contrastive approach significantly enhances prediction accuracy in both LLM and non-LLM settings, as demonstrated on drug-perturbation and generic perturbation datasets. AI

IMPACT Enhances LLM capabilities in scientific research by improving predictive accuracy for complex biological systems.

RANK_REASON Academic paper introducing a new method for LLM-based biological reasoning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• LLM
• Qwen3.5-9B

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →