Salsa Fresca: Angular Embeddings and Pre-Training for ML Attacks on Learning With Errors

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Stevens, Samuel, Wenger, Emily, Li, Cathy, Nolte, Niklas, Saxena, Eshika, Charton, François, Lauter, Kristin
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911769602031616
author Stevens, Samuel
Wenger, Emily
Li, Cathy
Nolte, Niklas
Saxena, Eshika
Charton, François
Lauter, Kristin
author_facet Stevens, Samuel
Wenger, Emily
Li, Cathy
Nolte, Niklas
Saxena, Eshika
Charton, François
Lauter, Kristin
contents Learning with Errors (LWE) is a hard math problem underlying recently standardized post-quantum cryptography (PQC) systems for key exchange and digital signatures. Prior work proposed new machine learning (ML)-based attacks on LWE problems with small, sparse secrets, but these attacks require millions of LWE samples to train on and take days to recover secrets. We propose three key methods -- better preprocessing, angular embeddings and model pre-training -- to improve these attacks, speeding up preprocessing by $25\times$ and improving model sample efficiency by $10\times$. We demonstrate for the first time that pre-training improves and reduces the cost of ML attacks on LWE. Our architecture improvements enable scaling to larger-dimension LWE problems: this work is the first instance of ML attacks recovering sparse binary secrets in dimension $n=1024$, the smallest dimension used in practice for homomorphic encryption applications of LWE where sparse binary secrets are proposed.
format Preprint
id arxiv_https___arxiv_org_abs_2402_01082
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Salsa Fresca: Angular Embeddings and Pre-Training for ML Attacks on Learning With Errors
Stevens, Samuel
Wenger, Emily
Li, Cathy
Nolte, Niklas
Saxena, Eshika
Charton, François
Lauter, Kristin
Cryptography and Security
Machine Learning
Learning with Errors (LWE) is a hard math problem underlying recently standardized post-quantum cryptography (PQC) systems for key exchange and digital signatures. Prior work proposed new machine learning (ML)-based attacks on LWE problems with small, sparse secrets, but these attacks require millions of LWE samples to train on and take days to recover secrets. We propose three key methods -- better preprocessing, angular embeddings and model pre-training -- to improve these attacks, speeding up preprocessing by $25\times$ and improving model sample efficiency by $10\times$. We demonstrate for the first time that pre-training improves and reduces the cost of ML attacks on LWE. Our architecture improvements enable scaling to larger-dimension LWE problems: this work is the first instance of ML attacks recovering sparse binary secrets in dimension $n=1024$, the smallest dimension used in practice for homomorphic encryption applications of LWE where sparse binary secrets are proposed.
title Salsa Fresca: Angular Embeddings and Pre-Training for ML Attacks on Learning With Errors
topic Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2402.01082