SALAD: Systematic Assessment of Machine Unlearning on LLM-Aided Hardware Design

TL;DR

This paper introduces SALAD, a systematic framework using machine unlearning to enhance data security in LLM-assisted hardware design by removing contaminated or sensitive information without full retraining.

Contribution

SALAD is the first comprehensive assessment applying machine unlearning to mitigate security threats in LLM-aided hardware design automation.

Findings

Machine unlearning effectively reduces data security risks.

Selective removal of sensitive data is feasible without full model retraining.

Case studies demonstrate improved security in hardware design workflows.

Abstract

Large Language Models (LLMs) offer transformative capabilities for hardware design automation, particularly in Verilog code generation. However, they also pose significant data security challenges, including Verilog evaluation data contamination, intellectual property (IP) design leakage, and the risk of malicious Verilog generation. We introduce SALAD, a comprehensive assessment that leverages machine unlearning to mitigate these threats. Our approach enables the selective removal of contaminated benchmarks, sensitive IP and design artifacts, or malicious code patterns from pre-trained LLMs, all without requiring full retraining. Through detailed case studies, we demonstrate how machine unlearning techniques effectively reduce data security risks in LLM-aided hardware design.