Entropy, Disagreement, and the Limits of Foundation Models in Genomics

Maxime Rochkoulets, Lovro Vrček, Mile Sikic

Confirmation: I have read and agree with the workshop's policy on behalf of myself and my co-authors.

Track: tiny / short paper (2-4 pages excluding references; extended abstract format)

Keywords: genomic foundation models, information theory, language modeling, fisher information

Abstract: Foundation models in genomics have shown mixed success compared to their counterparts in natural language processing. Yet, the reasons for their limited effectiveness remain poorly understood. In this work, we investigate the role of entropy as a fundamental factor limiting the capacities of such models to learn from their training data and develop foundational capabilities. We train ensembles of models on text and DNA sequences and analyze their predictions, static embeddings, and Fisher information flow. We show that the high entropy of genomic sequences—from the point of view of unseen token prediction—leads to near-uniform output distributions, disagreement across models, and unstable static embeddings, even for models that are matched in architecture, training and data. We then demonstrate that models trained on DNA concentrate Fisher information in embedding layers, seemingly failing to exploit inter-token relationships. Our results suggest that self-supervised training from sequences alone may not be suited for genomics, calling into question the assumptions underlying current methodologies for training genomic foundation models.

Anonymization: This submission has been anonymized for double-blind review via the removal of identifying information such as names, affiliations, and identifying URLs.

Submission Number: 59