paper-with-me

Papers

Gradient Projection For Continual Parameter-Efficient Tuning

2024-05-22 · Jingyang Qiao, Zhizhong Zhang, Xin Tan, Yanyun Qu, Wensheng Zhang, Zhi Han, Yuan Xie

Parameter-efficient tunings (PETs) have demonstrated impressive performance and promising perspectives in training large models, while they are still confronted with a common problem: the trade-off between learning new content and protecting old knowledge, leading to zero-shot generalization collapse, and cross-modal hallucination. In this paper, we reformulate Adapter, LoRA, Prefix-tuning, and Prompt-tuning from the perspective of gradient projection, and firstly propose a unified framework called Parameter Efficient Gradient Projection (PEGP). We introduce orthogonal gradient projection into different PET paradigms and theoretically demonstrate that the orthogonal condition for the gradient can effectively resist forgetting even for large-scale models. It therefore modifies the gradient towards the direction that has less impact on the old feature space, with less extra memory space and training time. We extensively evaluate our method with different backbones, including ViT and CLIP, on diverse datasets, and experiments comprehensively demonstrate its efficiency in reducing forgetting in class, online class, domain, task, and multi-modality continual settings. The project page is available at https://dmcv-ecnu-pegp.github.io/.

📄 PDF Abstract BibTeX arXiv:2405.13383

Code (0)

등록된 구현이 없습니다.

Tasks

Continual LearningHallucinationZero-shot Generalization

Methods 이 논문이 사용한 방법론

CLIP Contrastive Language-Image Pre-training (CLIP), consisting of a simplified version of ConVIRT trained from scratch, is an efficient method of image representation learning…
Adapter 설명 없음

Similar Papers 제목 키워드 기반

Continual Gradient Low-Rank Projection Fine-Tuning for LLMs

2025-07-03 · Chenxu Wang, Yilin Lyu, Zicheng Sun, Liping Jing

Continual fine-tuning of Large Language Models (LLMs) is hampered by the trade-off between efficiency and expressiveness. Low-Rank Adaptation (LoRA) offers efficiency but constrains the model's ability to learn new tasks…

Continual Learning

GEM-Style Constraints for PEFT with Dual Gradient Projection in LoRA

2026-01-05 · Brian Tekmen, Jason Yin, Qianqian Tong arxiv

Full fine-tuning of Large Language Models (LLMs) is computationally costly, motivating Continual Learning (CL) approaches that utilize parameter-efficient adapters. We revisit Gradient Episodic Memory (GEM) within the Lo…

Continual Learning

Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning

2026-05-09 · Binghang Lu, Zheyuan Deng, Runyu Zhang, Bing Hu 외 arxiv

A central challenge in continual learning for large language models (LLMs) is catastrophic forgetting, where adapting to new tasks can substantially degrade performance on previously learned ones. Existing projection-bas…

Continual Learning

Low-Rank Adapters Initialization via Gradient Surgery for Continual Learning

2026-05-12 · Joana Pasquali, Ramiro N. Barros, Arthur S. Bianchessi, Vinícius Conte Turani 외 arxiv

LoRA is widely adopted for continual fine-tuning of Large Language Models due to its parameter efficiency, modularity across tasks, and compatibility with replay strategies. However, LoRA-based continual learning remains…

Continual Learning

Restricted Orthogonal Gradient Projection for Continual Learning

2023-01-28 · Zeyuan Yang, Zonghan Yang, Peng Li, Yang Liu

Continual learning aims to avoid catastrophic forgetting and effectively leverage learned experiences to master new knowledge. Existing gradient projection approaches impose hard constraints on the optimization space for…

Continual LearningTransfer Learning