AI Paper Atlas

Research map / Federated learning and privacy / Privacy attacks and secure learning

Fed-FA: Theoretically Modeling Client Data Divergence for Federated Language Backdoor Defense

Zhiyuan Zhang, Deli Chen, Hao Zhou, Fandong Meng, Jie Zhou, Xu Sun

NeurIPS 2023 · Federated learning and privacy › Privacy attacks and secure learning · approach: Defending federated learning against poisoning · 1 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 f-divergence indicator instead of Euclidean indicator (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 58.50 -> 19.51Improved
use f-divergence indicator instead of Euclidean indicator, correlation with divergence (ablation)Fed-FA · dataset mixing ratio trialscorrelation: 0.9045 -> 0.9847Improved
add embedding Hessian reassignment (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 39.78 -> 19.51Improved
use synthesized dataset instead of labeled dataset for Hessian scales (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 20.06 -> 19.51No clear effect
use proportional reassignment instead of inverse reassignment (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 82.27 -> 19.51Improved
reassign Hessians only on embeddings instead of layer-wise (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 52.75 -> 19.51Improved
reassign Hessians only on embeddings instead of entire model (ablation)Fed-FA · Average over models, datasets, attacksASR (%): 42.03 -> 19.51Improved
Fed-FA versus Dim-Krum aggregationFed-FA · 3 models averageASR (%): 31.56 -> 19.51Improved
Fed-FA versus Dim-Krum on AgNewsFed-FA · AgNewsASR (%): 4.00 -> 3.43Improved
Fed-FA versus Dim-Krum on EP attackFed-FA · EP attackASR (%): 14.09 -> 12.13No clear effect
Fed-FA versus Dim-Krum on BadSent attackFed-FA · BadSent attackASR (%): 48.42 -> 28.30Improved
Fed-FA versus FedAvg clean accuracyFed-FA · 3 models averageclean ACC (%): 85.28 -> 85.70No clear effect

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