AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

SALSA: Attacking Lattice Cryptography with Transformers

Emily Wenger, Mingjie Chen, Francois Charton, Kristin Lauter

NeurIPS 2022 · Adversarial robustness and security › Attacks on text and graphs · approach: Membership inference and inversion · 10 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Universal Transformer with regular 8-layer transformer (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 22.5 -> 26.3Hurt
add copy-gate mechanism (ablation)SALSA UT · SALSA n=50, Hamming weight 3log2 samples to recover secret: 26.5 -> 22.6Improved
iterate shared layers 2/8 times instead of 4/4 (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 26.1 -> 23.5Improved
increase encoder dimension from 1024 to 3040 (decoder fixed at 512) (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 22.5 -> 19.7Improved
increase decoder dimension from 1024 to 1536 (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 23.9 -> 24.3Hurt
use small batch size (<50) (ablation)SALSA · SALSA n=50, Hamming weight 3samples neededImproved
reuse samples via linear combinations (K and times reused increase) (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 20.42 -> 15.405Improved
use higher input base (81 instead of 7) (ablation)SALSA · SALSA n=50, Hamming weight 3log2 samples to recover secret: 24.0 -> 22.1Improved
bound the coordinate range of a (a_i < alpha q) (ablation)SALSA · n=50 secret densitymax recoverable density: 0.08 -> 0.3Improved
increase error sigma up to 24 (ablation)SALSA · SALSA n=50, Hamming weight 3samples neededHurt
use distinguisher recovery instead of direct recovery only (ablation)SALSA · secret recoveryshare of recoveries: 55% direct, 18% distinguisher, 27% bothMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy