Metraipolis › archive › dc_2eeaa09661d7
Beneath the Diff: Diagnosing and Mitigating Algorithmic Mode Collapse in Code-Level Autonomous Research Loops — arXiv:2609.00077
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00077
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:41:02.841Z
- SHA-256 of the archived text
8320739a190e6c82e59d081f43bb65b4cce599c5d86ef9d1d5c72589af6bc5b1- Stable URL of this page
https://metraipolis.com/documento/dc_2eeaa09661d7
How to cite
Beneath the Diff: Diagnosing and Mitigating Algorithmic Mode Collapse in Code-Level Autonomous Research Loops — arXiv:2609.00077 — arXiv (en), https://arxiv.org/abs/2609.00077 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2eeaa09661d7 (Metraipolis archive dc_2eeaa09661d7, sha256 8320739a190e6c82e59d081f43bb65b4cce599c5d86ef9d1d5c72589af6bc5b1)
BibTeX
@misc{dc2eeaa09661d7,
title = {Beneath the Diff: Diagnosing and Mitigating Algorithmic Mode Collapse in Code-Level Autonomous Research Loops — arXiv:2609.00077},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00077},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2eeaa09661d7; sha256 8320739a190e6c82e59d081f43bb65b4cce599c5d86ef9d1d5c72589af6bc5b1},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00077. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Code-level autonomous research loops (ARLs) have recently emerged as a concrete object of study in automated machine learning research. In such loops, an LLM agent proposes modifications to an experimental training pipeline, executes the modified pipeline, and retains edits that improve a verifiable in-loop metric. Although executable metrics may appear to provide a reliable signal of progress, it remains unclear whether repeated metric-driven code editing leads to genuine improvements that generalize beyond the loop. We provide a systematic diagnosis of this question. Across various experiment settings, we identify a robust failure mode that we call \textbf{algorithmic mode collapse}. In this regime, surface-level edit diversity remains stable, but semantic and mechanism-level diversity collapse: the agent continues to edit different lines of code while repeatedly proposing the same kinds of algorithmic changes. This collapse is accompanied by a widening gap between in-loop metric gains and gains measured on independent held-out evaluations. We then propose Diversity-Aware Proposal Sampling (\textsc{DAPS}), a lightweight mitigation that combines category-coverage reweighting, persistent edit memory, and a validation gate. Under a three-tier protocol separating the in-loop metric, the audit metric read by the gate, and a blind metric no loop component ever accesses, \textsc{DAPS} reduces semantic-cluster decay of edits by $69.1\%$ and improves relative faithfulness by $83.7\%$ blind and $81.6\%$ audited, while preserving in-loop optimization speed. We provide the code in Github \href{ this https URL }{repository}.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2609.00077 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T06:33:38.784Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).