Summary of Investigating the Impact Of Quantization on Adversarial Robustness, by Qun Li et al.
Investigating the Impact of Quantization on Adversarial Robustness
by Qun Li, Yuan Meng, Chen Tang, Jiacheng Jiang, Zhi Wang
First submitted to arxiv on: 8 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates the impact of the quantization pipeline on the robustness of deep models against adversarial attacks. Quantization reduces the bit-width of models to improve runtime performance and storage efficiency, but its effects on model robustness have been understudied. The authors analyze the inconsistency in previous studies and find that it arises from differences in the pipelines used, particularly with regards to when and if robust optimization is performed. They demonstrate that quantization can affect model robustness and provide insights for deploying more secure and robust models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper looks at how a technique called quantization affects deep learning models. Quantization makes models faster and use less storage space, but it’s not clear what happens when the models are attacked by bad data. The authors studied previous studies on this topic and found that they got different results because of the way they did the quantization. They also found that quantization can make models less robust to attacks. |
Keywords
* Artificial intelligence * Deep learning * Optimization * Quantization
