Large Discovery Models: Empirically-grounded Model-Based Open-Ended Search
TL;DR - Large Discovery Model combines generative proposals with a continually updated Bayesian reward surrogate, enabling uncertainty-aware search across open-ended scientific design spaces. It outperforms LLM-only reflection and traditional statistical search in neural-network, antibody, and molecular optimization tasks.
- The surrogate estimates candidate performance and epistemic uncertainty from experimental observations.
- These estimates guide candidate generation, refinement, and selection while discovery memory updates recurrently.
- Reported gains include 2.4Ă— greater validation-BPB reduction, 18.2% lower binding energy, and over 60% relative improvement in molecular multi-objective performance.