Code Evolution Graphs: Understanding Large Language Model Driven Design of Algorithms Conference: Genetic and Evolutionary Computation Conference; July 14–18, 2025; Malaga, Spain Genetic and Evolutionary Computation Conference (GECCO ’25), July 14–18, 2025, Malaga, Spain Price: 15.00 DOI: 10.1145/36
Niki van Stein, Anna V. Kononova, Lars Kotthoff, Thomas Bäck · 2025
Read Code Evolution Graphs: Understanding Large Language Model Driven Design of Algorithms Conference: Genetic and Evolutionary Computation Conference; July 14–18, 2025; Malaga, Spain Genetic and Evolutionary Computation Conference (GECCO ’25), July 14–18, 2025, Malaga, Spain Price: 15.00 DOI: 10.1145/36 by Niki van Stein, Anna V. Kononova, Lars Kotthoff, Thomas Bäck (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-16668.txt.
Description
Niki van Stein and co-authors graph how AI-written algorithms evolve in a 2025 paper. Language models inside evolution loops draft code, test it, and draft again. Sometimes the loop stalls or yields weak code, and no one can see why from outside.
Code evolution graphs lay each draft and its children out as a family tree. Users can watch traits spread, die out, or grow tangled across prompts. Three benchmark problem classes supply the case studies. A clear trend stands out: repeated prompting grows code bulk, which can hurt speed. Seeing the process beats guessing at it, for builders of auto-design tools.
Catalog record
| CREATOR | Niki van Stein, Anna V. Kononova, Lars Kotthoff, Thomas Bäck |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-16668