AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

RS-Del: Edit Distance Robustness Certificates for Sequence Classifiers via Randomized Deletion

Zhuoqun Huang, Neil G Marchant, Keane Lucas, Lujo Bauer, Olga Ohrimenko, Benjamin I. P. Rubinstein

NeurIPS 2023 · Adversarial robustness and security › Certified and empirical robustness · approach: Randomized smoothing certification · 3 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
use randomized deletion smoothing (RS-Del) instead of randomized ablation (RS-Abn) with equal smoothing parametersMalConv malware detector · Sleipnir2 / VTFeedcertified accuracyImproved
smooth MalConv with RS-Del at p_del=99.5% versus non-smoothed modelMalConv · Sleipnir2clean accuracy: 2-3% dropHurt
vary decision threshold eta at p_del=99.5% (ablation)RS-Del · Sleipnir2median certified radius for malicious class: up to 582 bytesImproved
apply RS-Del smoothing instead of non-smoothed baseline against Disp attackMalConv · Sleipnir2 / VTFeedattack success rate: 20 percentage point decreaseImproved
apply RS-Del smoothing against six published attacksMalConv · Sleipnir2 / VTFeedattack success rateMixed
increase deletion probability p_del (ablation)RS-Del · Sleipnir2clean accuracy vs certified radiusMixed

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