Amortized Molecular Optimization via Group Relative Policy Optimization
Muhammad bin Javaid ∗, Hasham Hussain ∗, Ashima Khanna, Berke Kişin, Jonathan Pirnay, Alexander Mitsos, Dominik G. Grimm, Martin Grohe · 2026
Read Amortized Molecular Optimization via Group Relative Policy Optimization by Muhammad bin Javaid ∗, Hasham Hussain ∗, Ashima Khanna, Berke Kişin, Jonathan Pirnay, Alexander Mitsos, Dominik G. Grimm, Martin Grohe (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-12162.txt.
Description
Designing a better molecule usually means restarting a costly search for each new starting shape, which scales poorly when shapes or tests are dear. Muhammad bin Javaid led this work with seven co-authors, presenting AMORTIX, a graph transformer that learns the whole task once and then answers in a single pass, with no test-time oracle calls.
The hard part is that some starting shapes are far tougher than others, and plain training wobbles under that spread. The team steadies it by scaling rewards within groups that share one starting shape. Tests on kinase inhibitor design and a prodrug case show it beating rival learned and per-case methods, and a learned rule even transfers to unseen drug shapes.
Catalog record
| CREATOR | Muhammad bin Javaid ∗, Hasham Hussain ∗, Ashima Khanna, Berke Kişin, Jonathan Pirnay, Alexander Mitsos, Dominik G. Grimm, Martin Grohe |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-12162