Metraipolis › archive › dc_90ff691526bb
Dynamic Adaptation of the LLM Context for Generating Routines with Coupled Semantics — arXiv:2609.04570
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.04570
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T22:26:29.234Z
- SHA-256 of the archived text
2714606d127e171b60109c85deaa1082c77c7c9b156eceed59aeade05032d96b- Stable URL of this page
https://metraipolis.com/documento/dc_90ff691526bb
How to cite
Dynamic Adaptation of the LLM Context for Generating Routines with Coupled Semantics — arXiv:2609.04570 — arXiv (en), https://arxiv.org/abs/2609.04570 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_90ff691526bb (Metraipolis archive dc_90ff691526bb, sha256 2714606d127e171b60109c85deaa1082c77c7c9b156eceed59aeade05032d96b)
BibTeX
@misc{dc90ff691526bb,
title = {Dynamic Adaptation of the LLM Context for Generating Routines with Coupled Semantics — arXiv:2609.04570},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.04570},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_90ff691526bb; sha256 2714606d127e171b60109c85deaa1082c77c7c9b156eceed59aeade05032d96b},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.04570. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
LLM-based code generation fails when correctness depends on execution-dependent coupling: the meaning of one routine is defined by the runtime behavior of another, a relationship that cannot be resolved from textual descriptions alone. This limitation, which we call static binding, is not confined to explicitly coupled problems; it appears to varying degrees whenever correctness depends on joint execution behavior across components, from explicit cross-coupled optimizers to subtler joint constraints in packing, routing, and symbolic search. This paper proposes dynamic context adaptation, a sample-efficient validation-generation loop designed for this setting. A validation agent extracts structured diagnostic information from execution traces, providing gradient-like guidance to a generation agent that proposes multiple candidates per iteration. A knowledge graph derived from the problem description supplies semantic constraints to the generation agent. Simulated annealing selects among candidates to avoid greedy collapse. Our method outperforms zero-shot, Reflexion, and OpenEvolve on seven of eight problems at both 300 and 600 evaluations (p < 0.01), a regime where population-based search has not yet accumulated sufficient diversity to compete. Notably, on the primary motivating problem (cross-coupled optimization), our method also achieves the best score at 1000 evaluations, consistent with the hypothesis that structured execution feedback is most beneficial when correctness depends on runtime coupling. Ablation results confirm that structured execution feedback is the primary driver. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-07 in cs.AI). Categorie arXiv: Software Engineering (cs.SE) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.04570 — 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 2026-10-16T22:18:11.507Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-07?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).