An international eye on Swiss news

September 2026

An AI capable of doubt can optimize scientific discovery: A team at EPFL have developed a framework that trains Large Language Models to create AI systems that can find the best possible setup or "recipe" for a scientific experiment (LINK) — 1 September 2026

— Bojana Rankovic and Philippe Schwaller at EPFL's Laboratory of Artificial Chemical Intelligence have developed a method that combines an LLM with a a Gaussian process, a kind of 'doubt detector", the probabilistic model commonly used in Bayesian optimization." Rather than asking the LLM to choose experiments directly ("direct prompting"), GOLLuM trains it using the Gaussian process's way of scoring uncertainty and performance of each option. Because of this, as the LLM learns from past experiments, it adjusts how it "organizes" the search space, the "master list" of every possible choice for an experiment.

— The researchers evaluated GOLLuM across 23 benchmark tasks covering organic synthesis, analytical and process chemistry, materials and catalysis, and molecular property optimization. Each optimization run began with ten low-performing observations, and the team used the same GOLLuM configuration across all benchmark tasks rather than tuning it separately for each problem, showing its versatility. Across the 23 benchmarks, GOLLuM ranked first on average, finding high-performing experimental conditions more consistently than Bayesian optimization with expert-designed descriptors.