ADAPT for Science: A Multi-Agent System for Self-Evolving Scientific Optimization

Scientific progress often depends on finding a small number of useful solutions in vast and constrained design spaces. Scientific automation and embodied intelligence have made experiments and simulations easier to execute. Yet deciding what to try next remains difficult when evaluations are costly, feasible solutions are sparse, and parameters are strongly coupled. Recent AI for AI methods can automatically improve algorithms and optimization procedures. However, general purpose approaches are not designed around scientific priors, physical constraints, and limited evaluation budgets. We introduce ADAPT for Science, a multi-agent framework that treats scientific optimization as two coupled decisions: where to search and how to search. ADAPT uses historical designs to identify promising and diverse initial conditions, progressively concentrates computation on favorable search directions, and evolves the optimization procedure from execution feedback. We evaluate ADAPT on the design of gravitational wave detectors using a differentiable physics simulator and eight unseen detector topologies. Under matched scientific evaluation budgets, ADAPT achieves a mean loss of $0.203\pm 0.106$, reducing loss by $55.1\%$ relative to the best classical optimizer and by $62.4\%$ relative to the best general purpose agentic evolution baseline. ADAPT achieves the lowest loss on all unseen topologies. ADAPT provides a step toward scientific optimization systems that continuously learn from prior experience and their own search outcomes.