Research map / Label-efficient and robust learning / Domain adaptation and generalization

MTL-LoRA: Low-Rank Adaptation for Multi-Task Learning

AAAI 2025 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Mixture and distillation for domain shift · 17 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes MTL-LoRA, a LoRA variant for multi-task learning that adds task-specific low-rank transformations to capture task differences; however, the provided full text is only the AAAI 2025 author template, so method details and results are unavailable.

Problem

LoRA in multi-task learning projects sparse high-dimensional features from different tasks into one shared dense low-rank space, which obscures task distinctions.

Approach

Based on the abstract, MTL-LoRA is a parameter-efficient adaptation for multi-task learning that aims to preserve task distinctions in the low-rank space; the specific mechanism could not be verified because the provided text is only the AAAI formatting template.

Limitations

• Full text provided is the AAAI template rather than the paper body, so method and results are not verifiable

Experimental setup

Builds on: LoRA

Closest papers by meaning

• MTLoRA: Low-Rank Adaptation Approach for Efficient Multi-Task Learning (CVPR 2024 · similarity 0.90)
• Beyond Higher Rank: Token-wise Input-Output Projections for Efficient Low-Rank Adaptation (NeurIPS 2025 · similarity 0.89)
• Uni-LoRA: One Vector is All You Need (NeurIPS 2025 · similarity 0.89)
• MeteoRA: Multiple-tasks Embedded LoRA for Large Language Models (ICLR 2025 · similarity 0.89)
• AuroRA: Breaking Low-Rank Bottleneck of LoRA with Nonlinear Mapping (NeurIPS 2025 · similarity 0.89)
• Beyond LoRA vs. Full Fine-Tuning: Gradient-Guided Optimizer Routing for LLM Adaptation (NeurIPS 2026 · similarity 0.89)