AI Paper Atlas

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

NatADiff: Adversarial Boundary Guidance for Natural Adversarial Diffusion

Max Collins, Jordan Vice, Tim French, Ajmal Saeed Mian

ICLR 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Diffusion-based adversarial purification · 4 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 NatADiff instead of constrained attacks PGD/AARN-50 surrogate · ImageNet avg transfer ASRASR (%): 17.6 / 18.4 -> 68.2Improved
use NatADiff (untargeted) instead of AdvClass (untargeted)RN-50 surrogate · ImageNet avg ASRASR (%): 45.7 -> 68.2Improved
use untargeted similarity targeting vs random targetsNatADiff RN-50 surrogate · ImageNet avg ASRASR (%): 56.8 -> 68.2Improved
targeted vs untargeted NatADiff on ViT-H surrogateNatADiff · ImageNet avg ASRASR (%): untargeted 69.7 vs targeted 73.2 (targeted IS 15.3 vs 31.9)Mixed
add classifier augmentations (ablation)NatADiff RN-50 surrogate, targeted · ImageNet avg ASRASR (%): 48.7 -> 56.8Improved
add classifier augmentations (ablation)NatADiff RN-50 surrogate, targeted · ImageNet image qualityFID-Val: 56.7 -> 66.5Hurt
add classifier augmentations (ablation)NatADiff RN-50 surrogate, targeted · ImageNet-A alignmentFID-A: 81.7 -> 77.3Improved
increase boundary guidance mu from 0.0 to 0.5 (ablation)NatADiff RN-50 surrogate · ImageNet avg ASRASR (%): 52.1 -> 69.5Improved
increase boundary guidance mu (ablation)NatADiff RN-50 surrogate · FID-AFID-A: lowest 77.3 at mu=0.2; 80.1 at mu=0.5Mixed
change intersection prompt template (ablation)NatADiff · avg ASR across 8 templatesASR (%): varies by less than 7.7%No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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