AI Paper Atlas

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

Exposing and Defending the Achilles' Heel of Video Mixture-of-Experts

Songping Wang, Qinglong Liu, Yueming Lyu, Ning Li, Ziwen He, Caifeng Shan

ICLR 2026 · Adversarial robustness and security › Certified and empirical robustness · approach: Robust overfitting and regularization · 5 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
add joint router-and-expert attack (J-TLGA) instead of router-only TLGA-RVideo MoE · UCF-101 eps=8/255robust accuracy (%): TLGA-R 18.13 -> J-TLGA 4.95 (stronger attack lowers accuracy)Improved
add Lipschitz regularity and temporal adaptive step to PGD (TLA-M vs PGD)Video MoE attack · UCF-101 eps=8/255robust accuracy (%): 22.09 -> 15.05 (lower is stronger attack)Improved
add Lipschitz term to router attack gradient (TLGA-R vs PGD-R)Video MoE attack · UCF-101 eps=8/255robust accuracy (%): 42.20 -> 18.13 (lower is stronger attack)Improved
add temporal adaptive step to expert attack (TLA-E vs PGD-E)Video MoE attack · UCF-101 eps=8/255robust accuracy (%): 31.98 -> 28.11 (lower is stronger attack)Improved
add joint component adversarial training (J-TLAT) over TLATAT-M MoE · UCF-101 PGD eps=8/255robust accuracy (%): 30.22 -> 36.37Improved
add joint component adversarial training (J-TLAT) over TLATAT-M MoE · UCF-101 J-TLGA eps=8/255robust accuracy (%): 9.78 -> 33.96Improved
add J-TLAT over TLATAT-M MoE · UCF-101 TLGA-R eps=8/255robust accuracy (%): 26.37 -> 51.87Improved
add J-TLAT over TLATMoE · UCF-101 router Lipschitz constantLips-R: 3.500 -> 0.823Improved
use J-TLAT instead of AAT-MMoE · UCF-101 J-TLGA eps=8/255robust accuracy (%): 0.00 -> 33.96Improved
use TLA as inner attack instead of PGD in J-TLAT (ablation)J-TLAT · UCF-101 PGD evalrobust accuracy (%): 33.63 (a 5.5% gain over conventional PGD)Improved
use TLA as inner attack in J-TLAT (ablation)J-TLAT · UCF-101 J-TLGA evalrobust accuracy (%): 30.33Improved
use J-TLAT regularization instead of JS, COS, KL, NR, GR (ablation)J-TLAT · UCF-101 FGSM/PGD/J-TLGArobust accuracy (%): 36.59, 33.63, 30.33Improved

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