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

French

Le code généré par les grands modèles linguistiques passe souvent tous les tests fonctionnels tout en restant incertain. Ces hallucinations de sécurité sont courantes, écrites avec une grande confiance et répétées sur les échantillons, de sorte que ni l'incertitude du générateur, ni les analystes statiques ne les capturent de manière fiable. Nous introduisons SPLASH, qui est guidée par une nouvelle mesure appelée pression d'audit. Il contraste les probabilités de journaux forcées par l'enseignant pour le programme généré sous un examen de sécurité et un examen de style. Aggregé aux déclarations et aux programmes, et fondé sur un modèle probabilistique d'éditions concurrentes, la pression d'audit permet à SPLASH de choisir le programme sécurisé parmi les programmes candid

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