Open-Ended Task Discovery via Bayesian Optimization
For practitioners of Bayesian optimization in scientific workflows, GSR addresses the overlooked problem of task uncertainty by enabling automatic task discovery, though the approach is incremental.
GSR introduces an open-ended Bayesian optimization framework that dynamically generates and refines tasks, achieving logarithmic regret overhead relative to single-task BO and outperforming LLM-based optimizers in product development, chemical synthesis, algorithm analysis, and patent repurposing.
When applying Bayesian optimization (BO) to scientific workflow, a major yet often overlooked source of uncertainty is the task itself -- namely, what to optimize and how to evaluate it -- which can evolve as evidence accumulates. We introduce Generate-Select-Refine (GSR), a open-ended BO framework that alternates between task generation and task optimization. Starting from a user-provided seed task, GSR generates new tasks in a coarse-to-fine manner while a task-acquisition function schedules optimization. Asymptotically, it concentrates evaluations on the best task, incurring only logarithmic regret overhead relative to single-task BO. We apply GSR to new product development, chemical synthesis scaling, algorithm analysis, and patent repurposing, where it outperforms existing LLM-based optimizers.