AI Paper Atlas

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

Harnessing Hyperbolic Geometry for Harmful Prompt Detection and Sanitization

Igor Maljkovic, Maria Rosaria Briglia, Iacopo Masi, Antonio Emanuele Cinà, Fabio Roli

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Vision-language model robustness · 0 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 hyperbolic HSVDD on HySAC embeddings instead of Euclidean SVDD on CLIP (ablation)HyPE · ViSU testF1: 0.66 -> 0.98Improved
use hyperbolic HSVDD on HySAC embeddings instead of Euclidean SVDD on CLIP (ablation)HyPE · ViSU testprecision: 0.08 -> 0.98Improved
increase nu from 0.01 to about 0.0325 (ablation)HSVDD · ViSU validationF1: peak F1 around 0.0325Improved
increase nu from 0.1 toward 1 (ablation)HSVDD · ViSU validationbenign accuracyHurt
use HyPE instead of Detoxifydetection · ViSUF1: 0.40 -> 0.98Improved
use HyPE instead of NSFW-Classifierdetection · MMAF1: 0.75 -> 0.95Improved
use HyPE instead of GuardT2Idetection · adv-MMAF1: 0.19 -> 0.96Improved
use HyPE instead of NSFW-Classifier on SneakyPromptdetection · SneakyPromptF1: 0.78 -> 0.78No clear effect
use Word Removal instead of Thesaurus+LLM sanitizationHyPS · ViSUfraction reclassified benign (%): 65% -> 85%Improved
use Thesaurus+LLM instead of Word Removal sanitizationHyPS · ViSUsemantic similarity: 0.82 and 0.87Improved
increase adaptive attack lambda toward 1HyPE · ViSU adaptive attack (1,000 prompts)detectionHurt
apply HyPE filtering against Concept Combination AttackHySAC encoder in SD · ViSU safe promptsunsafe image occurrences: reduced to zeroImproved

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