Latent jepa helps predict chemical outcomes from hidden reasoning steps
Latent JEPA: Abstract Future Prediction for Latent Reasoning in Chemistry
Machine LearningComputation and Language
Summary
Predicting chemical reactions and molecular changes is tricky because it involves many steps and possibilities. The authors made a new method called Latent JEPA that trains a computer model to anticipate important future results without explaining every tiny step out loud. This helps the model think ahead and connect what it 'thinks' internally to actual chemical outcomes better. When tested, this approach improved the model’s ability to suggest useful changes to molecules and predict reactions.
What this means in practice
- •For chemical informatics teams: Improve molecular editing and reaction predictions by integrating latent future outcome prediction into chemical reasoning models.
- •For computational chemistry software developers: Develop tools for molecular optimization that anticipate future molecular states without explicit intermediate steps, boosting efficiency.
Authors
Xinjian Zhao, Yaoyao Xu, Xuemin Chen, Xiaozhuang Song, Tianshu Yu
Abstract
Large language models offer a promising foundation for chemical reasoning, bringing together chemical knowledge and multistep problem solving. Chemical intuition can provide an initial sense of plausible outcomes before the details of a solution are fully worked out. Inspired by how such expectations complement explicit analysis, we study how continuous latent thoughts can be trained to anticipate informative aspects of future solutions without verbalizing every intermediate step. We introduce Latent JEPA, a framework that combines autoregressive learning with joint-embedding prediction of one or more future views. For chemical reasoning, we develop textual and molecular prediction objectives that connect latent thoughts to both subsequent reasoning and molecular outcomes. Experiments on ChemCoTBench show gains in molecular optimization and on several editing and reaction metrics. Representation analyses show that future prediction makes latent thoughts more informative about molecular outcomes and strengthens their correspondence with chemical structure. These findings support abstract future prediction as a learning principle for connecting continuous latent reasoning with scientific outcomes.