Design Principles for Controlled Non-Equilibrium Self-Assembly as Computation

Molecular self-assembly provides a potential physical pathway to computation, where collective organization encodes solutions to complex problems. We investigate methodologies for uncovering general design principles that prepare and constrain self-assembling systems so they can serve as fair and reliable computational media. Rather than attempting to solve specific problems or enforce particular outcomes, we aim to define transferable physical rules that allow systems with continual state changes to reach structured steady states capable of meaningful exploration of configuration space. By focusing on general constraints rather than finely tuned interactions, we seek to avoid over-encoding individual problem instances and instead identify conditions that make diverse computations physically realizable. Through analytical and statistical-mechanical reasoning, this work explores how broad constraints can make non-equilibrium assembly stable, interpretable, and computationally expressive—laying the groundwork for transforming molecular systems into tunable, rule-based platforms for large-scale, physics-driven computation through self-organization.