AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

SG-LoRA: Semantic-guided LoRA Parameters Generation

Miaoge Li, Yang Chen, Zhijie Rao, Can Jiang, Kang Wei, Jingcai Guo

CVPR 2026 · Large language models › Instruction and fine-tuning · approach: Low-rank adaptation variants · 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
generate LoRA with SG-LoRA instead of Top-k weighted mergingTop-k LoRA Weighted · MS-COCO retrievalI2T R@1: 71.55 -> 74.31Improved
generate LoRA with SG-LoRA instead of Oracle fine-tuned LoRASG-LoRA · MS-COCO retrievalT2I R@1: 53.10 -> 54.42Improved
generate LoRA with SG-LoRA instead of Oracle fine-tuned LoRASG-LoRA · OxfordPets retrievalI2T R@1: 55.84 -> 57.15Improved
weight top-k experts with softmax similarity instead of equal weights (ablation)Top-k LoRA merging · MS-COCO retrievalI2T R@1: 70.70 -> 71.55Improved
select semantic top-k experts instead of uniformly averaging all (Model Soups) (ablation)Model Soups · MS-COCO retrievalT2I R@1: 47.38 -> 49.85Improved
train the generator on MS-COCO and apply to OxfordPetsSG-LoRA · OxfordPets retrievalI2T R@1: 57.15 -> 55.41Hurt
train generator on OxfordPets and test on MS-COCOSG-LoRA · MS-COCO retrievalI2T R@1: 74.31 -> 70.81Hurt
use textual semantic prior instead of visual prior (ablation)SG-LoRA · MS-COCO retrievalI2T R@1: 73.16 -> 74.31Improved
use textual semantic prior instead of visual prior (ablation)SG-LoRA · CIFAR-100 superclassaccuracy: 73.83 -> 77.50Improved
use detailed description listing classes instead of generic photo prompt (ablation)SG-LoRA · CIFAR-100 superclassaccuracy: 75.77 -> 77.50Improved
add MS-COCO Cat expert to the repository (ablation)SG-LoRA · OxfordPets Egyptian MauI2T R@1: 36.08 -> 37.11Improved
add MS-COCO Cat expert to the repository (ablation)SG-LoRA · OxfordPets PersianI2T R@5: 80.00 -> 79.65Hurt

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