Saltar al contenido
← volver

Poison Valleys: Diagnosing Path-Dependent Failure in LLM Code Repair

arXiv cs.AI · arxiv.org Publicado:

arXiv:2610.21514v1 Announce Type: cross Abstract: When code-repair agents fail, final tests reveal that a patch is wrong but not which earlier commitment caused it or which software obligation failed. Poison Valleys traces this path in 1,844 model requests on two synthetic Python cases: PersistQueue adds atomic batch claiming to a disk-backed queue, while Dijkstra adds source-sensitive filtering to shortest-path search. We define plausible but incorrect software mechanisms, place each in a mode…

Leer original →

Más de arXiv cs.AI