Keumseo Ryum | Searcharxiv

arXiv2026

CT-Merging: Consensus Directions and Task-Level Scaling for LoRA Adapter Merging

LoRA adapters provide an efficient way to specialize a pretrained model for many downstream tasks, but deploying one adapter per task requires adapter storage and task selection at inference time. Model merging addresses this issue by combining independently trained adapters into one multi-task adapter. Recent SVD-based LoRA merging methods mainly focus on constructing shared or task specific directions, while the coefficients assigned to the final directions are often directly from the original task SVD. On a fixed merged basis, inherited coefficients preserve component order with high rank correlation, yet their magnitudes differ substantially from the coefficients induced by the task updates. To address this mismatch, we propose CT-Merging, a LoRA-aware merging algorithm that estimates consensus directions from average task subspace projectors and assigns task-level RMS coefficient scales in the final update. CT-Merging uses repeated support across task SVD subspaces to construct the common basis, while reducing reliance on rank wise SVD magnitudes after direction construction. On the DC-Merge CLIP adapter benchmark, CT-Merging achieves superior average normalized accuracy compared to state-of-the-art merging methods and further improves over DC-Merge by 2.56 points on ViT-B/32 and 1.51 points on ViT-L/14 KnoTS-trained checkpoints.

cs.LG↗

arXiv2025

Resource-Aware Aggregation and Sparsification in Heterogeneous Ensemble Federated Learning

Federated learning (FL) enables distributed training with private client data, but its convergence is hindered by system heterogeneity under realistic communication scenarios. Most FL schemes addressing system heterogeneity utilize global pruning or ensemble distillation, yet often overlook typical constraints required for communication efficiency. Meanwhile, deep ensembles can aggregate predictions from individually trained models to improve performance, but current ensemble-based FL methods fall short in fully capturing diversity of model predictions. In this work, we propose \textbf{SHEFL}, a global ensemble-based FL framework suited for clients with diverse computational capacities. We allocate different numbers of global models to clients based on their available resources. We introduce a novel aggregation scheme that mitigates the training bias between clients and dynamically adjusts the sparsification ratio across clients to reduce the computational burden of training deep ensembles. Extensive experiments demonstrate that our method effectively addresses computational heterogeneity, significantly improving accuracy and stability compared to existing approaches.

cs.LG↗