Summary of Algebraic Adversarial Attacks on Integrated Gradients, by Lachlan Simpson et al.
Algebraic Adversarial Attacks on Integrated Gradients
by Lachlan Simpson, Federico Costanza, Kyle Millar, Adriel Cheng, Cheng-Chew Lim, Hong Gunn Chew
First submitted to arxiv on: 23 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Group Theory (math.GR)
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 A novel approach to generating adversarial attacks on explainability models, specifically path methods like integrated gradients, is proposed in this research paper. The authors introduce algebraic adversarial examples as a way to create mathematically tractable attacks that can be used to evaluate the robustness of attribution methods used in safety-critical systems. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re trying to understand why a self-driving car made a certain decision, but someone is trying to trick the system by creating fake explanations. This paper explores ways to defend against these “adversarial attacks” on explainability models, like integrated gradients. The authors develop a new method for generating these attacks that’s easier to work with mathematically. |
