Results 91 to 100 of about 9,649,421 (248)

Low-Rank Adaptation of Pre-Trained Large Vision Models for Improved Lung Nodule Malignancy Classification

IEEE Open Journal of Engineering in Medicine and Biology
Goal: This paper investigates using Low-Rank Adaptation (LoRA) to adapt large vision models (LVMs) pretrained with self-supervised learning (SSL) for lung nodule malignancy classification.
Benjamin P. Veasey, Amir A. Amini
doaj   +1 more source

The role of digital media in the home learning environment and associations with children's early language and literacy skills

British Journal of Educational Technology, EarlyView.
Abstract Young children increasingly engage with digital media tools for purposes of education, entertainment and communication. This study examines children's digital media activities within the context of their larger home learning environment (HLE).
Burcu H. Tatar, Lori E. Skibbe, Ryan P. Bowles, Hope K. Gerde   +3 more
wiley   +1 more source

SaLoRA: Safety-Alignment Preserved Low-Rank Adaptation [PDF]

As advancements in large language models (LLMs) continue and the demand for personalized models increases, parameter-efficient fine-tuning (PEFT) methods (e.g., LoRA) will become essential due to their efficiency in reducing computation costs.
Zhang, Yang, Si, Wai Man, Wang, Yisen, Li, Mingjie, Backes, Michael   +4 more
core   +2 more sources

Parameter-Efficient Adaptation of Qwen2.5 for Aspect-Based Sentiment Analysis Using Low-Rank Adaptation and Parameter-Efficient Fine-Tuning

Engineering Proceedings
Aspect-based sentiment analysis (ABSA) plays a vital role in deriving fine-grained sentiment from textual content. As large language models (LLMs) are increasingly adopted for automated data annotation in natural language processing (NLP), concerns have ...
Pei Ying Lim, Chuk Fong Ho, Chi Wee Tan
doaj   +1 more source

Computational Limits of Low-Rank Adaptation (LoRA) Fine-Tuning for Transformer Models

Accepted at ICLR 2025.
Jerry Yao-Chieh Hu, Maojiang Su, En-Jui Kuo, Zhao Song 0002, Han Liu 0001   +4 more
openaire   +3 more sources

Interlaboratory performance testing on EPIC v2.0 CNS tumor profiling demonstrates high reproducibility of tumor classification but reveals the need for harmonized copy number variation reporting

Brain Pathology, EarlyView.
Using the Infinium MethylationEPIC v2.0 array and Heidelberg Brain Tumor Classifier v12.8, 24 international laboratories achieved highly reproducible CNS tumor classification (97.9% correct; median β‐correlation r = 0.99), while copy number variation interpretation showed substantial interlaboratory variability, highlighting the need for harmonized CNV
Katrin Mauch‐Mücke, Christin Siewert, Lora Dimitrova, Sandra Pohl, Brendan Osberg, Annika K. Wefers, Julia Lorenz, Tanja Rothhammer‐Hampl, Till Acker, Alonso Barrantes‐Freer, Jean Boutonnat, Bérengère Dadone‐Montaudié, Claire Delbridge, Wolfgang Dietmaier, Sebastian Dintner, Hildegard Dohmen, Matthias Dottermusch, Daniel Erny, Katrin B. M. Frauenknecht, Tobias Goschzik, Patrick N. Harter, Daniela Hirsch, Yanghao Hou, Theo F. J. Kraus, Maggie Lee, Friederike Liesche‐Starnecker, Anne Mc Leer, Linea Melchior, Michael Mittelbronn, Wolf C. Müller, Shalima S. Nair, Julia E. Neumann, Pitt Niehusmann, Nicole Pfarr, Torsten Pietsch, Marco Prinz, Markus Rechsteiner, Regina Reimann, Laveniya Satgunaseelan, Veronica Skarpeteig, Anne J. Skjulsvik, Karl Sotlar, Isabelle Vanden Bempt, Erik Vassella, Katharina J. Weber, Gabriel Wasinger, Andreas Waha, Karen Zwaenepoel, Felix Sahm, Markus J. Riemenschneider, David Capper   +50 more
wiley   +1 more source

Fine-tuning CLIP in spectral space for multimodal sentiment analysis

Complex & Intelligent Systems
Recent advances in Parameter-Efficient Fine-Tuning (PEFT) have enabled effective adaptation of large-scale pre-trained models, particularly vision–language architectures such as CLIP, with substantially reduced training cost.
Ju Qin, Yuntao Sun
doaj   +1 more source

LoRA-Fine-Tuned Latent Diffusion for High-Fidelity Digitization of Classic Mongolian Patterns

Applied Sciences
Mongolian patterns represent an important component of Mongolian cultural heritage, characterized by their dual structure of geometric symmetry and dynamic ornamental motifs.
Jiatong Liu, Yue Huang
doaj   +1 more source

Under the radar: a longitudinal exploration of mental health among children and adolescents experiencing parental and caregiver death during the COVID‐19 pandemic in South Africa

Child and Adolescent Mental Health, EarlyView.
Background Death of a caregiver during childhood can have profound influences on child wellbeing and later trajectories. Globally, child and adolescent mental health is an increasing area of concern with widespread negative implications. These data provide the first comprehensive exploration of the mental health of children experiencing COVID‐19 ...
Kathryn Steventon Roberts, Christina A. Laurenzi, Tatenda Mawoyo, Stefani Du Toit, Mark Tomlinson, Lucie Cluver, Lorraine Sherr   +6 more
wiley   +1 more source

SD-LoRA: Scalable Decoupled Low-Rank Adaptation for Class Incremental Learning

Continual Learning (CL) with foundation models has recently emerged as a promising paradigm to exploit abundant knowledge acquired during pre-training for tackling sequential tasks. However, existing prompt-based and Low-Rank Adaptation-based (LoRA-based) methods often require expanding a prompt/LoRA pool or retaining samples of previous tasks, which ...
Yichen Wu, Hongming Piao, Long-Kai Huang, Renzhen Wang, Wanhua Li 0001, Hanspeter Pfister, Deyu Meng, Kede Ma, Ying Wei 0001   +8 more
openaire   +3 more sources