Algorithmic algorithm development with LLMs: A Case Study on LLM-Usage for Contraction Order Optimization in Tensor Networks
For researchers in algorithm development and tensor networks, this work demonstrates the potential and limitations of using LLMs for automated algorithm improvement, but the results are incremental and domain-specific.
This paper explores LLM-based algorithm development for tensor network contraction order optimization, showing that verifier-guided evolutionary coding agents can improve algorithms but emphasizing the continued need for human oversight in evaluation and interpretation.
We consider LLM-based algorithm development through a case study on contractionorder optimisation for tensor networks with OpenEvolve. We pay particular attention to the choice of the LLM as well as design choices such as evaluation metric and test instances. Our results highlight both the promise of verifier-guided evolutionary coding agents for algorithm development/improvement and the continuing importance of evaluation, validation, and interpretation -- and corresponding challenges -- by the human scientist.