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

Static analysis is effective for discovering software vulnerabilities but notoriously suffers from incomplete source--sink specifications and excessive false positives (FPs). We present textscAdaTaint, an LLM-driven taint analysis framework that adaptively infers source/sink specifications and filters spurious alerts through neuro-symbolic reasoning. Unlike LLM-only detectors, textscAdaTaint grounds model suggestions in program facts and constraint validation, ensuring both adaptability and determinism. We evaluate textscAdaTaint on Juliet 1.3, SV-COMP-style C benchmarks, and three large real-world projects. Results show that textscAdaTaint reduces false positives by 43.7% on average and improves recall by 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.