GLoRA (Parameter-efficient fine-tuning (LoRA family)): current frontier — recent, not yet superseded in the knowledge base. 0 paper(s) critique it, 0 beat it on benchmarks — not ranked as a superseded baseline. Sub-problem: cluster led by FFA-LoRA. Newer alternatives in the same sub-problem include Shift-Dependent Asymmetry, GLoRA, HiLoRA, FedRot-LoRA, ADF-LoRA.

Parameter-efficient fine-tuning (LoRA family)

GLoRA

Beyond Factor Aggregation: Gauge-Aware Low-Rank Server Representations for Federated LoRA

On the frontierfirst seen May 7, 2026

Current frontier — recent, not yet superseded in the knowledge base

0 papers critique it · 0 beat it on benchmarks

What to use instead

Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.

Shift-Dependent AsymmetryShift-Dependent Asymmetry: Orthogonal Inverse Low-Rank Adaptation for Federated Medical Segmentation
Jun 7, 2026
GLoRABeyond Factor Aggregation: Gauge-Aware Low-Rank Server Representations for Federated LoRA
May 7, 2026
HiLoRAHiLoRA: Hierarchical Low-Rank Adaptation for Personalized Federated Learning
Mar 3, 2026
FedRot-LoRAFedRot-LoRA: Mitigating Rotational Misalignment in Federated LoRA
Feb 27, 2026
ADF-LoRAADF-LoRA: Alternating Low-Rank Aggregation for Decentralized Federated Fine-Tuning
Nov 23, 2025
compute-efficient continual pre-training with LoRALow-Resource Dialect Adaptation of Large Language Models: A French Dialect Case-Study
Oct 26, 2025
personalized federated fine-tuning method with orthogonal LoRA adaptersPersonalized Federated Fine-Tuning of Vision Foundation Models for Healthcare
Oct 14, 2025
FLoRA-NACommunication-Efficient and Accurate Approach for Aggregation in Federated Low-Rank Adaptation
Sep 30, 2025