Unleashing the Power of Generative Model in Recovering Variable Names from Stripped Binary
Xiangzhe Xu
Network and Distributed System Security (NDSS) Symposium 2025 · Day 3 · Binary Analysis
In this compelling talk at the NDSS Symposium, Xiangzhe Xu from Purdue University presented groundbreaking research on recovering variable names from stripped binaries using generative code models. The work introduces a novel context-aware fine-tuning technique and a preference optimization method to align model generations with developer naming conventions. This advancement is crucial for democratizing reverse engineering, making complex binary programs more accessible to a broader audience—from security experts to ordinary users without a computer science background.
AI review
Solid academic research that makes a genuine contribution to the binary analysis problem — generative models over classifiers for symbol recovery is the right call, the context-aware fine-tuning trick is clever, and the preference optimization to fight long-tail bias shows the authors actually thought through the failure modes rather than just benchmarking happy-path cases. Not a world-shaker, but this is real work that will get cited and built upon.