Discover Latent Biology with Mechanistic Reinforcement Learning

Uncover novel biological insights with our GReinSS approach.


Beyond Representations, Directly Learning Hidden Biological States

Machine learning (AI) shows great promise in theory but often fails to deliver sufficient impact on real-world biological problems. While existing machine learning techniques learn artificial continuous representations of the data, at ReinBio, we focus on directly modeling discrete mechanistic biological states of interest through biologically grounded generative models.

Publications

S Ivanovic, G Liu, M El-Kebir

ICML 2026

Logo for the International Conference on Machine Learning (ICML), featuring a stylized human head with multicolored pixels and the text 'ICML' and 'International Conference on Machine Learning'.

S Ivanovic, M El-Kebir

Genome Biology 2025

Logo for the journal Genome Biology.
Logo for the journal Genome Biology.

S Ivanovic, M El-Kebir

Genome Research 2023

Logo for the journal Genome Research.