LLM-Driven Adaptive Source-Sink Identification and False Positive Mitigation for Static Analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Lin, Shiyin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912690187796480
author Lin, Shiyin
author_facet Lin, Shiyin
contents Static analysis is effective for discovering software vulnerabilities but notoriously suffers from incomplete source--sink specifications and excessive false positives (FPs). We present \textsc{AdaTaint}, an LLM-driven taint analysis framework that adaptively infers source/sink specifications and filters spurious alerts through neuro-symbolic reasoning. Unlike LLM-only detectors, \textsc{AdaTaint} grounds model suggestions in program facts and constraint validation, ensuring both adaptability and determinism. We evaluate \textsc{AdaTaint} on Juliet 1.3, SV-COMP-style C benchmarks, and three large real-world projects. Results show that \textsc{AdaTaint} reduces false positives by \textbf{43.7\%} on average and improves recall by \textbf{11.2\%} compared to state-of-the-art baselines (CodeQL, Joern, and LLM-only pipelines), while maintaining competitive runtime overhead. These findings demonstrate that combining LLM inference with symbolic validation offers a practical path toward more accurate and reliable static vulnerability analysis.
format Preprint
id arxiv_https___arxiv_org_abs_2511_04023
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LLM-Driven Adaptive Source-Sink Identification and False Positive Mitigation for Static Analysis
Lin, Shiyin
Software Engineering
Cryptography and Security
Static analysis is effective for discovering software vulnerabilities but notoriously suffers from incomplete source--sink specifications and excessive false positives (FPs). We present \textsc{AdaTaint}, an LLM-driven taint analysis framework that adaptively infers source/sink specifications and filters spurious alerts through neuro-symbolic reasoning. Unlike LLM-only detectors, \textsc{AdaTaint} grounds model suggestions in program facts and constraint validation, ensuring both adaptability and determinism. We evaluate \textsc{AdaTaint} on Juliet 1.3, SV-COMP-style C benchmarks, and three large real-world projects. Results show that \textsc{AdaTaint} reduces false positives by \textbf{43.7\%} on average and improves recall by \textbf{11.2\%} compared to state-of-the-art baselines (CodeQL, Joern, and LLM-only pipelines), while maintaining competitive runtime overhead. These findings demonstrate that combining LLM inference with symbolic validation offers a practical path toward more accurate and reliable static vulnerability analysis.
title LLM-Driven Adaptive Source-Sink Identification and False Positive Mitigation for Static Analysis
topic Software Engineering
Cryptography and Security
url https://arxiv.org/abs/2511.04023