Newest papers

SPLASH: Contrastive Audit Pressure for Detecting and Localizing Security Hallucinations in Generated Code

Mirza Masfiqur Rahman, Ajwad Akil, Imtiaz Karim, Ashish Kundu, Ramana Rao Kompella, Elisa Bertino · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26001 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

Estas alucinaciones de seguridad son comunes, escritas con alta confianza y repetidas a través de muestras, por lo que ni la incertidumbre propia del generador ni los analizadores estáticos las capturan de manera fiable. Introducimos SPLASH, que está guiado por una nueva medida llamada presión de auditoría. Contrasta las probabilidades de registro forzadas por el profesor para el programa generado bajo una revisión de seguridad y una revisión de estilo. Agregado a declaraciones y programas, y basado en un modelo probabilístico de ediciones competidoras, la presión de auditoría permite a SPLASH elegir el programa seguro entre los programas candidatos de un modelo y localizar la declaración vulnerable, sin formación del modelo, reglas específicas de CWE o acceso interno al modelo generador. En 63.684 programas generados por 12 modelos para 899 tareas,

Machine translation (Meta M2M100); the English original follows.

Abstract

Code generated by large language models often passes every functional test while remaining insecure. These security hallucinations are common, written with high confidence, and repeated across samples, so neither the generator's own uncertainty nor static analyzers reliably catch them. We introduce SPLASH, which is guided by a new measure called audit pressure. It contrasts teacher-forced log-probabilities for the generated program under a security review and a style review. Aggregated to statements and programs, and grounded in a probabilistic model of competing edits, audit pressure allows SPLASH to choose the secure program among a model's candidate programs and locate the vulnerable statement, with no model training, CWE-specific rules, or internal access to the generating model. On 63,684 programs generated by 12 models for 899 tasks, SPLASH picks secure code in 74% of cases, places the vulnerable statement first in 58% of insecure programs, and chooses better than a frontier LLM judge.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Code (set theory), Audit, Probabilistic logic, Computer security, Measure (data warehouse)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange