Psi-Evolve: Autonomously Illuminating the Space of Problems

Using modern LLMs and large amounts of resources, AI systems have become capable of solving increasingly complex existing scientific problems. However, to achieve truly autonomous scientific discovery, AI systems must also be able to propose diverse problems that go beyond readily available human knowledge. This is inherently challenging for LLMs, which have been trained on human sources and verifiable objectives, and may lack the experience and intuition to find interesting new problems with potential downstream utility. In this work, we introduce Psi-Evolve ($\Psi$-Evolve), new framework leveraging evolutionary computation to bridge this gap. $\Psi$-Evolve combines carefully designed descriptors, mutations, and objectives to find diverse problems whose solutions are inherently generalizable. Empirically, we show that $\Psi$-Evolve autonomously discovers archives of new problems that span well beyond the space covered by readily available human archives and whose solutions can zero-shot transfer across a given scientific domain. Furthermore, we also show that $\Psi$-Evolve's archives can be used in conjunction with prior AI systems, accelerating and cutting the cost of LLM-driven solution discovery itself. Together with this submission, we open-source our full implementation and the data used in its development, collected from hundreds of human problems, which we hope will support future research toward expanding the scope of autonomous scientific inquiry and discovery.