AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

LoRA.rar: Learning to Merge LoRAs via Hypernetworks for Subject-Style Conditioned Image Generation

Donald Shenaj, Ondrej Bohdal, Mete Ozay, Pietro Zanuttigh, Umberto Michieli

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Multi-subject personalization · 13 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
replace test-time optimization (ZipLoRA) with a pre-trained hypernetwork predicting merging coefficientsLoRA merging · MARS2 test pairs (SDXL)MARS2 average case: 0.58 -> 0.71Improved
replace per-pair optimization with a single forward pass of the hypernetworkLoRA merging · Footprinttime to predict merging coefficients: 158s -> 0.037sImproved
apply LoRA.rar on a lightweight backboneLoRA.rar · KOALA 700mMARS2 average case: 0.51 -> 0.56Improved
hypernetwork-merge query and output LoRAs, average key and value (ablation)LoRA.rar · Validation pairsMARS2 average case: 0.23 -> 0.48 (all four vs Query, Output)Improved
use hypernetwork on query, value and output only (ablation)LoRA.rar · Validation pairsMARS2 best case: 0.92 -> 0.33Hurt
use MARS2 instead of CLIP-I or DINO to track human preferenceEvaluation metric · Direct merge human studycorrelation with human score: CLIP-I 0.08, DINO -0.01, MARS2 0.76Improved
train hypernetwork on narrow splits with no category overlap (ablation)LoRA.rar · New splitsMARS2Mixed
LoRA.rar on unseen furniture subjects and cyberpunk styleLoRA.rar · New conceptsMARS2: 0.6 -> 0.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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