Summary of A Survey on Concept-based Approaches For Model Improvement, by Avani Gupta et al.
A survey on Concept-based Approaches For Model Improvement
by Avani Gupta, P J Narayanan
First submitted to arxiv on: 21 Mar 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG)
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 presents a systematic review and taxonomy of concept representation methods and automatic concept discovery algorithms for Deep Neural Networks (DNNs) in computer vision applications. It highlights the shift from improving DNN performance to developing more interpretable models, particularly in the field of eXplainable Artificial Intelligence (XAI). The authors explain how concepts enable detecting biases, correlations, and clever-hans by providing simple human-understandable terms for model decisions. They also discuss concept-based model improvement methods, which can enhance interpretability and generalization. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about making computer models more understandable to humans. Right now, these models are really good at doing certain tasks, but we don’t know why they’re making the decisions they are. To fix this, researchers have been working on ways to explain how the model is thinking, using something called “concepts”. These concepts help us understand when the model might be biased or getting confused. The paper takes a close look at different methods for finding and using these concepts in computer vision tasks. |
Keywords
* Artificial intelligence * Generalization
