ExpProof : Operationalizing Explanations for Confidential Models with ZKPs
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910974596874240 |
|---|---|
| author | Yadav, Chhavi Laufer, Evan Monroe Boneh, Dan Chaudhuri, Kamalika |
| author_facet | Yadav, Chhavi Laufer, Evan Monroe Boneh, Dan Chaudhuri, Kamalika |
| contents | In principle, explanations are intended as a way to increase trust in machine learning models and are often obligated by regulations. However, many circumstances where these are demanded are adversarial in nature, meaning the involved parties have misaligned interests and are incentivized to manipulate explanations for their purpose. As a result, explainability methods fail to be operational in such settings despite the demand \cite{bordt2022post}. In this paper, we take a step towards operationalizing explanations in adversarial scenarios with Zero-Knowledge Proofs (ZKPs), a cryptographic primitive. Specifically we explore ZKP-amenable versions of the popular explainability algorithm LIME and evaluate their performance on Neural Networks and Random Forests. Our code is publicly available at https://github.com/emlaufer/ExpProof. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_03773 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ExpProof : Operationalizing Explanations for Confidential Models with ZKPs Yadav, Chhavi Laufer, Evan Monroe Boneh, Dan Chaudhuri, Kamalika Machine Learning Artificial Intelligence Cryptography and Security In principle, explanations are intended as a way to increase trust in machine learning models and are often obligated by regulations. However, many circumstances where these are demanded are adversarial in nature, meaning the involved parties have misaligned interests and are incentivized to manipulate explanations for their purpose. As a result, explainability methods fail to be operational in such settings despite the demand \cite{bordt2022post}. In this paper, we take a step towards operationalizing explanations in adversarial scenarios with Zero-Knowledge Proofs (ZKPs), a cryptographic primitive. Specifically we explore ZKP-amenable versions of the popular explainability algorithm LIME and evaluate their performance on Neural Networks and Random Forests. Our code is publicly available at https://github.com/emlaufer/ExpProof. |
| title | ExpProof : Operationalizing Explanations for Confidential Models with ZKPs |
| topic | Machine Learning Artificial Intelligence Cryptography and Security |
| url | https://arxiv.org/abs/2502.03773 |