Influence-based Attributions can be Manipulated
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_ | 1866909338135691264 |
|---|---|
| author | Yadav, Chhavi Wu, Ruihan Chaudhuri, Kamalika |
| author_facet | Yadav, Chhavi Wu, Ruihan Chaudhuri, Kamalika |
| contents | Influence Functions are a standard tool for attributing predictions to training data in a principled manner and are widely used in applications such as data valuation and fairness. In this work, we present realistic incentives to manipulate influence-based attributions and investigate whether these attributions can be \textit{systematically} tampered by an adversary. We show that this is indeed possible for logistic regression models trained on ResNet feature embeddings and standard tabular fairness datasets and provide efficient attacks with backward-friendly implementations. Our work raises questions on the reliability of influence-based attributions in adversarial circumstances. Code is available at : \url{https://github.com/infinite-pursuits/influence-based-attributions-can-be-manipulated} |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_05208 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Influence-based Attributions can be Manipulated Yadav, Chhavi Wu, Ruihan Chaudhuri, Kamalika Machine Learning Artificial Intelligence Influence Functions are a standard tool for attributing predictions to training data in a principled manner and are widely used in applications such as data valuation and fairness. In this work, we present realistic incentives to manipulate influence-based attributions and investigate whether these attributions can be \textit{systematically} tampered by an adversary. We show that this is indeed possible for logistic regression models trained on ResNet feature embeddings and standard tabular fairness datasets and provide efficient attacks with backward-friendly implementations. Our work raises questions on the reliability of influence-based attributions in adversarial circumstances. Code is available at : \url{https://github.com/infinite-pursuits/influence-based-attributions-can-be-manipulated} |
| title | Influence-based Attributions can be Manipulated |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2409.05208 |