Results 141 to 150 of about 227,550 (285)

Instruction-Tuned Decoder-Only Large Language Models for Efficient Extreme Summarization on Consumer-Grade GPUs

open access: yesAlgorithms
Extreme summarization generates very short summaries, typically a single sentence, answering the question “What is the document about?”. Although large language models perform well in text generation, fine-tuning them for summarization often requires ...
Attia Fathalla Elatiky   +3 more
doaj   +1 more source

Interconnected Porous Hydrogels with Tunable Anisotropy Through Aqueous Emulsion Bioprinting

open access: yesAdvanced Functional Materials, EarlyView.
A 3D bioprintable microporous bioink is developed using an aqueous two‐phase system (ATPS) composed of extracellular matrix (ECM) mimetic biopolymers. The ATPS bioink enables the fabrication of interconnected porous architectures with up to 70% porosity, supporting long‐term cell viability and 3D cell alignment, enabling a simultaneous generation of ...
Hugo Edgar‐Vilar   +4 more
wiley   +1 more source

Automatic Scoring of Arabic Essays: A Parameter-Efficient Approach for Grammatical Assessment

open access: yesIEEE Access
Automatic essay scoring (AES) utilizes computational methods to evaluate written essays, offering an efficient approach to grading in various educational contexts. Essays can have multiple scoring criteria.
Somaia Mahmoud, Emad Nabil, Marwan Torki
doaj   +1 more source

Pore Size Effects of Mesoporous N‐Doped Carbon Nanospheres as Advanced Support Material on the Activity of Molybdenum Sulfide Catalysts for the Hydrogen Evolution Reaction

open access: yesAdvanced Functional Materials, EarlyView.
By tuning the pore size of mesoporous N‐doped carbon (MPNC) nanospheres as support material for molybdenum sulfide, the electrochemical activity of the composite material for the hydrogen evolution reaction can be optimized. An ideal MPNC pore size of 60 nm allows a high number of molybdenum sulfide active sites while maintaining efficient proton and ...
Niklas Ortlieb   +3 more
wiley   +1 more source

DC-LoRA: Domain correlation low-rank adaptation for domain incremental learning

open access: yesHigh-Confidence Computing
Continual learning, characterized by the sequential acquisition of multiple tasks, has emerged as a prominent challenge in deep learning. During the process of continual learning, deep neural networks experience a phenomenon known as catastrophic ...
Lin Li   +4 more
doaj   +1 more source

Electrochemically Driven Dissipative Growth of Affinity Hydrogels for Bioresponsive Interfaces

open access: yesAdvanced Functional Materials, EarlyView.
Voltage pulses drive the growth and reinforcement of hydrogel films under dissipative conditions. This biocompatible strategy enables efficient integration of affinity ligands into the hydrogel matrix, enhancing the selective capture of growth factors and allowing precise temporal control over their release, making them well‐suited as adaptive ...
Roberto Baretta, Marco Frasconi
wiley   +1 more source

DualDyConvNet: Dual-Stream Dynamic Convolution Network via Parameter-Efficient Fine-Tuning for Predicting Motor Prognosis in Subacute Stroke

open access: yesIEEE Transactions on Neural Systems and Rehabilitation Engineering
Stroke is a significant impediment on a global scale, with the prognosis for motor ability contingent on initial rehabilitation and the severity of the injury.
Yunjeong Jang   +7 more
doaj   +1 more source

Memba: Membrane-driven Parameter-Efficient Fine-Tuning for Mamba

open access: yes
State Space Models (SSMs) have emerged as powerful alternatives to attention-based Transformers, with Mamba demonstrating impressive efficiency and scalability. As these models grow increasingly larger, the need for Parameter-Efficient Fine-Tuning (PEFT) methods becomes critical to adapt pre-trained Mamba to downstream tasks without prohibitive ...
Lee, Donghyun   +4 more
openaire   +2 more sources

Bayesian Parameter-Efficient Fine-Tuning for Overcoming Catastrophic Forgetting

open access: yesIEEE/ACM Transactions on Audio, Speech, and Language Processing
We are motivated primarily by the adaptation of text-to-speech synthesis models; however we argue that more generic parameter-efficient fine-tuning (PEFT) is an appropriate framework to do such adaptation. Nevertheless, catastrophic forgetting remains an issue with PEFT, damaging the pre-trained model's inherent capabilities.
Haolin Chen, Philip N. Garner
openaire   +2 more sources

Viscoelasticity‐Induced Controllable Periodic Meso‐Textures of Liquid Crystal Polymers in Additive Manufacturing

open access: yesAdvanced Functional Materials, EarlyView.
Viscoelasticity‐driven instabilities are harnessed to create tunable, periodic textures in 3D‐printed liquid crystalline polymers. This study illustrates how processing parameters control these spontaneous meso‐scale patterns. These unique structural architectures unlock new possibilities for functional devices, ranging from photonic components to ...
Miaomiao Zou   +17 more
wiley   +1 more source

Home - About - Disclaimer - Privacy