AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

SafeR-CLIP: Mitigating NSFW Content in Vision-Language Models While Preserving Pre-Trained Knowledge

Adeel Yousaf, Joseph Fioresi, James Beetham, Amrit Singh Bedi, Mubarak Shah

AAAI 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Feature calibration for unseen-class detection

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
add proximity-aware redirection and relative redirection loss over Safe-CLIPSafeR-CLIP · zero-shot classification, 11 datasetsaccuracy (%): 52.2 -> 60.2Improved
add relative redirection loss (ablation)Safe-CLIP · ViSU / zero-shotT*->V R@1 and ZS avg: 14.5 -> 28.0; ZS 52.2 -> 56.4Improved
add proximity-based alignment (ablation)SafeR-CLIP · zero-shot avgaccuracy (%): 56.4 -> 58.6Improved
add progressive training (ablation)SafeR-CLIP · zero-shot avgaccuracy (%): 58.6 -> 60.2Improved
apply proximity alignment to uni-modal losses as well as cross-modal (ablation)SafeR-CLIP · ViSU V*->TR@1: 18.7 -> 24.6Improved
average top-2 or top-3 safe targets instead of top-1 (ablation)SafeR-CLIP · ViSU T*->V and V*->TR@1: T*->V 27.9 -> 26.6 (top-2), 25.5 (top-3)Hurt
use JINA-CLIP instead of CLIP for caption matching (ablation)SafeR-CLIP · zero-shot avgaccuracy (%): 60.2 -> 59.0Hurt
use SafeR-CLIP on real NSFW imagesSafeR-CLIP · SMID V->T retrieval% NSFW retrieved: 14.2 -> 3.1Improved
replace Safe-CLIP with SafeR-CLIP in Stable DiffusionSD v1.4 · I2Pavg NSFW score: 16.1 -> 16.0No clear effect
apply SafeR-CLIP to SigLIPSigLIP-SO400M · NSFWCapsT*->V R@1: 46.7 -> 72.1; ZS 71.3 -> 76.6Improved
use SafeR-CLIP encoder in LLaVA-OneVision-7BLLaVA-OneVision-7B · MMBench, NudeNetbenign task score vs NSFW %: MMBench 81.8 -> 78.0; NudeNet 19.6 -> 9.4Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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