AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

TaskIT: Memory-Efficient Fine-Tuning of Multi-LoRA LLMs via Cross-Task Importance Transfer

Cheng Fang, Zimu Zhou, Ke Ma, Bin Guo

CVPR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: LoRA variants and theory

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 task similarity assessment instead of uniform task similarities (ablation)TaskIT · image tasksaccuracy: 75.3 -> 80.4Improved
add module importance predictor instead of uniform importances (ablation)TaskIT · image tasksaccuracy: 74.1 -> 80.4Improved
add block-based memory predictor instead of uniform memory model (ablation)TaskIT · image taskspeak memory (GB): 9.9 -> 7.8Improved
use DP scheduler instead of greedy (ablation)TaskIT · image tasksaccuracy: 77.5 -> 80.4Improved
fine-tune with TaskIT instead of best Zero-FT (AdapterFusion)multi-LoRA Janus-Pro · cross-modal tasksavg accuracy: +10.4%Improved
TaskIT instead of best non-LoRA baseline UniPTmulti-LoRA Janus-Pro · cross-modal tasksmemory and accuracy: -13.8% memory, +1.5% accuracyImproved
TaskIT instead of best LoRA baseline AutoLoRAmulti-LoRA Janus-Pro · cross-modal tasksmemory and accuracy: -45.5% memory, -0.3% accuracyMixed
TaskIT versus UniPT on uni-modal image tasksmulti-LoRA Janus-Pro · VTABavg accuracy: +2.1% accuracy, -11.4% memoryImproved
smaller memory budget from 12 GB to 8 GB in continual learning (ablation)TaskIT · continual task sequenceaccuracy gain vs LoRAHurt
fall back to uniform importance when no task has similarity above 0.35TaskIT · audio captioning (t9)importance transferNo clear effect

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