AI Paper Atlas

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

Oblivionis: A Lightweight Learning and Unlearning Framework for Federated Large Language Models

Fuyao Zhang, Xinyu Yan, Tiantong Wu, Wenjie Li, Tianxiang Chen, Yang Cao, Ran Yan, Longtao Huang, Wei Yang Bryan Lim, Qiang Yang

AAAI 2026 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated prompt and LoRA tuning · 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 FedAdagrad instead of FedAvgSimNPO unlearning (Llama-3.2-1B) · TOFU Split99Forget Truth Ratio: 0.65 -> 0.74Improved
use FedAdagrad instead of FedAvgSimNPO unlearning (Llama-3.2-1B) · TOFU Split99Model Utility: 0.46 -> 0.42Hurt
use FedAvgM instead of FedAvgSimNPO unlearning (Llama-3.2-1B) · TOFU Split99Model Utility: 0.46 -> 1.8e-4Hurt
use FedProx instead of FedAvgGradDiff unlearning (Llama-2-7B) · MUSE NewsUtility Preserved: 0.34 -> 0.52Improved
use FedProx instead of FedAvgGradDiff unlearning (Llama-2-7B) · MUSE NewsNo Verbatim Memorization: NVM 0.39 -> 0.52Hurt
scale base model from Llama-3.2-1B to 3BFedAvg + Retrain · TOFU Split99Model Utility: 0.51 -> 0.59Improved
use federated learning instead of local trainingGradAscent · TOFU Split99 (Llama-3.2-1B)Model Utility: 0.00 -> 0.46Improved
use federated learning instead of local trainingSimNPO · TOFU Split99 (Llama-3.2-1B)Model Utility: 0.35 -> 0.46Improved
apply unlearning algorithms after federated fine-tuningFedAvg (Llama-3.2-1B) · TOFU Split99Forget Truth Ratio: 0.49 -> 0.65Improved
apply unlearning after federated fine-tuningFedAvg (Llama-3.2-1B) · TOFU Split99Model Utility: 0.50 -> 0.46Hurt

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