AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

Rethinking LoRA for Privacy-Preserving Federated Learning in Large Models

Jin Liu, Yinbin Miao, Ning Xi, Junkang Liu

ICLR 2026 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated prompt and LoRA tuning · 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
add local alternating updates of A and B (ablation)DP-LoRA · Tiny-ImageNet (Swin-B, eps=3)test accuracy (%): 30.64 -> 53.07Improved
add local alternating updates of A and B (ablation)DP-LoRA · QNLI (RoBERTa-Base, eps=3)test accuracy (%): 86.31 -> 88.92Improved
add Gaussian low-pass filter (ablation)LA-LoRA without filter · Tiny-ImageNet (Swin-B, eps=3)test accuracy (%): 53.07 -> 61.97Improved
add Gaussian low-pass filter (ablation)DP-LoRA · Tiny-ImageNet (Swin-B, eps=3)test accuracy (%): 30.64 -> 49.85Improved
add Gaussian low-pass filter (ablation)DP-LoRA · SST-2 (RoBERTa-Base, eps=3)test accuracy (%): 92.36 -> 92.52No clear effect
replace Doppler filter with Gaussian filter (ablation)LA-LoRA · Tiny-ImageNet (Swin-B, eps=3)test accuracy (%): +4.61%Improved
vary local alternating schedule (1,2,5 steps per factor, A-first or B-first) (ablation)LA-LoRA · CIFAR-100/Tiny-ImageNet (Swin-B)test accuracy (%): best-worst gap below 0.7%No clear effect
increase LoRA rank 8 -> 32 (ablation)LA-LoRA · CIFAR-100 (Swin-B, eps=1)test accuracy (%): 72.88 -> 75.34Improved
use LA-LoRA instead of RoLoRARoLoRA · Tiny-ImageNet (Swin-B, eps=1)test accuracy (%): 43.85 -> 60.68Improved
use LA-LoRA instead of RoLoRARoLoRA · MNLI (RoBERTa-Base, eps=3)test accuracy (%): 82.12 -> 82.99Improved
use LA-LoRA on Llama-2-7B instead of RoLoRARoLoRA · GLUE avg (Llama-2-7B, eps=1)average accuracy (%): 88.66 -> 89.36Improved
use local alternating updatesDP-LoRA · Max Hessian eigenvalue (Swin-B CIFAR-100, eps=1)max Hessian eigenvalue: 101.62 -> 64.77Improved

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