Agentic generation of verifiable rules for deterministic, self-expanding reaction classification
2026-07-01 • Artificial Intelligence
Artificial IntelligenceComputation and Language
AI summaryⓘ
The authors developed an automated system that uses large language models to learn and create reaction rules from a huge collection of chemical reactions found in patents. Unlike traditional methods that rely on fixed, manually made rules, their system can generate and check its own rules, covering many more types of reactions without human help. This allows the system to classify almost all new reactions accurately and adapt to new chemistry it hasn't seen before. Essentially, they created a constantly improving chemical reaction database powered by AI.
computer-assisted synthesis planninglarge language modelsreaction ruleschemical reactionsclassificationpatent reactionsreactivity databasemachine learningsymbolic systemstaxonomy expansion
Authors
Daniel Armstrong, Maarten Dobbelaere, Valentas Olikauskas, Helena Avila, Octavian Susanu, Jérôme Waser, Philippe Schwaller
Abstract
Computer-assisted synthesis planning breaks target molecules into accessible precursors using large libraries of reaction rules that assign each transformation a deterministic, interpretable label. But chemistry is long-tailed, making manual encoding intractable, and existing tools rely on fixed rulesets that cannot adapt to new chemistries. Here we present a fully automated pipeline in which a multi-agent framework of large language models (LLMs) classifies reactions and writes the rules themselves across 665,901 US patent reactions, generating each rule under a verification loop that tests it against the corpus. It expands a standard taxonomy from 68 to 14,073 classes without human curation. With a lightweight fingerprint classifier, it classifies 97.7\% of unseen reactions, matching a leading proprietary classifier while resolving chemistry more finely and extending on demand to chemistry outside its training distribution. The result is a living reactivity database and a general route to turning generative models into reliable, self-expanding symbolic systems.