The SaTML '24 CNN Interpretability Competition: New Innovations for Concept-Level Interpretability
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909160361164800 |
|---|---|
| author | Casper, Stephen Yun, Jieun Baek, Joonhyuk Jung, Yeseong Kim, Minhwan Kwon, Kiwan Park, Saerom Moore, Hayden Shriver, David Connor, Marissa Grimes, Keltin Nicolson, Angus Tagade, Arush Rumbelow, Jessica Nguyen, Hieu Minh Hadfield-Menell, Dylan |
| author_facet | Casper, Stephen Yun, Jieun Baek, Joonhyuk Jung, Yeseong Kim, Minhwan Kwon, Kiwan Park, Saerom Moore, Hayden Shriver, David Connor, Marissa Grimes, Keltin Nicolson, Angus Tagade, Arush Rumbelow, Jessica Nguyen, Hieu Minh Hadfield-Menell, Dylan |
| contents | Interpretability techniques are valuable for helping humans understand and oversee AI systems. The SaTML 2024 CNN Interpretability Competition solicited novel methods for studying convolutional neural networks (CNNs) at the ImageNet scale. The objective of the competition was to help human crowd-workers identify trojans in CNNs. This report showcases the methods and results of four featured competition entries. It remains challenging to help humans reliably diagnose trojans via interpretability tools. However, the competition's entries have contributed new techniques and set a new record on the benchmark from Casper et al., 2023. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_02949 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | The SaTML '24 CNN Interpretability Competition: New Innovations for Concept-Level Interpretability Casper, Stephen Yun, Jieun Baek, Joonhyuk Jung, Yeseong Kim, Minhwan Kwon, Kiwan Park, Saerom Moore, Hayden Shriver, David Connor, Marissa Grimes, Keltin Nicolson, Angus Tagade, Arush Rumbelow, Jessica Nguyen, Hieu Minh Hadfield-Menell, Dylan Machine Learning Artificial Intelligence Interpretability techniques are valuable for helping humans understand and oversee AI systems. The SaTML 2024 CNN Interpretability Competition solicited novel methods for studying convolutional neural networks (CNNs) at the ImageNet scale. The objective of the competition was to help human crowd-workers identify trojans in CNNs. This report showcases the methods and results of four featured competition entries. It remains challenging to help humans reliably diagnose trojans via interpretability tools. However, the competition's entries have contributed new techniques and set a new record on the benchmark from Casper et al., 2023. |
| title | The SaTML '24 CNN Interpretability Competition: New Innovations for Concept-Level Interpretability |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2404.02949 |