LLM Driven Design of Continuous Optimization Problems with Controllable High-level Properties

TL;DR

This paper presents a method using large language models within an evolutionary framework to generate diverse continuous optimization problems with specific high-level landscape properties, enhancing benchmarking capabilities.

Contribution

It introduces the LLaMEA framework that guides LLMs to create problem instances with targeted features, expanding the diversity of benchmark test suites.

Findings

Generated problems exhibit desired structural traits.

The method increases landscape diversity and reduces redundancy.

The problem library is broad, interpretable, and reproducible.

Abstract

Benchmarking in continuous black-box optimisation is hindered by the limited structural diversity of existing test suites such as BBOB. We explore whether large language models embedded in an evolutionary loop can be used to design optimisation problems with clearly defined high-level landscape characteristics. Using the LLaMEA framework, we guide an LLM to generate problem code from natural-language descriptions of target properties, including multimodality, separability, basin-size homogeneity, search-space homogeneity and globallocal optima contrast. Inside the loop we score candidates through ELA-based property predictors. We introduce an ELA-space fitness-sharing mechanism that increases population diversity and steers the generator away from redundant landscapes. A complementary basin-of-attraction analysis, statistical testing and visual inspection, verifies that many of the generated functions indeed exhibit the intended structural traits. In addition, a t-SNE embedding shows that they expand the BBOB instance space rather than forming an unrelated cluster. The resulting library provides a broad, interpretable, and reproducible set of benchmark problems for landscape analysis and downstream tasks such as automated algorithm selection.