Bolek: A Multimodal Language Model for Molecular Reasoning

Frederic Grabowski, Jacek Szczerbiński, Maciej Jaśkowski, Kalina Jasinska-Kobus, Pawel Dabrowski-Tumanski, Tomasz Jetka, Bartosz Topolski

Keywords: molecular language models, multimodal learning, molecular reasoning, drug discovery, grounded reasoning

TL;DR: Bolek is a compact molecular language model that aligns Morgan fingerprints with feature-grounded reasoning to make drug-discovery property predictions more accurate and auditable.

Abstract:

AI models and platforms for molecular science underpin high-stakes applications in drug discovery, yet the systems delivering them are largely opaque: they expose either a score and a binary answer with no rationale, or fluent prose rarely anchored in molecular structure.
We introduce Bolek, a compact multimodal language model that injects a molecular embedding into an instruction-tuned decoder to predict molecular properties and explain them in auditable terms.
Bolek outperforms the chemistry-specialist TxGemma-9B-Chat on 13 of 15 downstream binary classification tasks and exceeds it on five unseen TDC endpoints, while grounding its reasoning with concrete numerical and structural descriptors.
Together, these results suggest that targeted modality injection paired with reasoning supervision tied to verifiable features can match domain-specialist systems at a fraction of the parameter count while producing auditable explanations for downstream decisions.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 58