AI Paper Atlas

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

Not All Clients Are Equal: Collaborative Model Personalization on Heterogeneous Multi-Modal Clients

Minhyuk Seo, Taeheon Kim, Hankook Lee, Jonghyun Choi, Tinne Tuytelaars

ICLR 2026 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Personalization with global-local trade-off · 0 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 Co-LoRA (shareable P,Q modules) over vanilla homogeneous averaging (ablation)FedMosaic · DRAKE-Dynamic OthersAlast: 47.63 -> 51.33Improved
add RELA relevance-guided aggregation on top of Co-LoRA (ablation)FedMosaic · DRAKE-Dynamic SelfAlast: 66.99 -> 67.86Improved
add RELA on top of Co-LoRA (ablation)FedMosaic · Fed-Scope SelfAlast: 28.47 -> 30.58Improved
add RELA on top of Co-LoRA (ablation)FedMosaic · DRAKE-Dynamic OthersAlast: 51.33 -> 51.16No clear effect
use FedMosaic instead of local-only SFTFedMosaic · DRAKE-Dynamic SelfAlast: 65.79 -> 67.86Improved
use standard PFL methods (DITTO, FedDAT, PerAda) instead of local SFTPFL baselines · DRAKE-Dynamic SelfAlast: 65.79 -> 58.47 (FedDAT)Hurt
use FedMosaic instead of SFT on HFLB (single-image VQA)FedMosaic · HFLB-Dynamic SelfAlast: 79.99 -> 80.80Improved
use FedMosaic across cross-family Llama/Qwen clientsFedMosaic · DRAKE-Dynamic cross-family SelfAlast: 68.60 -> 70.67Improved
use FedMosaic with 52 heterogeneous LLM clientsFedMosaic · Fed-LLM-Large SelfAlast: 18.03 -> 20.87Improved
use FedMosaic for complex multi-image task clients (3B models)FedMosaic · DRAKE-Dynamic client 5 SelfAlast: 62.00 -> 68.87Improved
use FedMosaic for simple single-image VQA clientsFedMosaic · DRAKE-Dynamic client 3 SelfAlast: 66.63 -> 66.71No clear effect
vary ratio of 1B and 3B clients (ablation)FedMosaic · DRAKE-Dynamic SelfAlast: +1.45 to +2.68 over SFTImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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