A retrosynthesis model that explicitly learns strategic bond-disconnection reasoning via reinforcement learning with a round-trip accuracy reward.
arXiv · March 16, 2026 · 2603.12666
Why it matters
Moving beyond simple sequence-to-sequence reactant prediction, this system emulates chemical expert strategies, resulting in more feasible and diverse proposals for organic synthesis.
From the abstract
Retrosynthesis prediction is a core task in organic synthesis that aims to predict reactants for a given product molecule. Traditionally, chemists select a plausible bond disconnection and derive corresponding reactants, which is time-consuming and requires substantial expertise. While recent advancements in molecular large language models (LLMs) have made progress, many methods either predict reactants without strategic reasoning or conduct only a generic product analysis, rather than reason ex