Applying an Agentic Coding Tool for Improving Published Algorithm Implementations
For researchers and peer reviewers, this work demonstrates a practical AI-assisted method for improving published code, but the results are incremental and lack concrete performance numbers.
The authors present a two-stage pipeline using an LLM and Claude Code to improve published algorithm implementations, reporting improvements in all eleven experiments within a single working day each.
We present a two-stage pipeline for AI-assisted improvement of published algorithm implementations. In the first stage, a large language model with research capabilities identifies recently published algorithms satisfying explicit experimental criteria. In the second stage, Claude Code is given a prompt to reproduce the reported baseline and then iterate an improvement process. We apply this pipeline to published algorithm implementations spanning multiple research domains. Claude Code reported that all eleven experiments yielded improvements. Each improvement could be achieved within a single working day. We analyse the human contributions that remain indispensable, including selecting the target, verifying experimental validity, assessing novelty and impact, providing computational resources, and writing with appropriate AI-use disclosure. Finally, we discuss implications for peer review and academic publishing.