Results 51 to 60 of about 33,388 (259)

SVFT: Parameter-Efficient Fine-Tuning with Singular Vectors

Advances in Neural Information Processing Systems 37
Popular parameter-efficient fine-tuning (PEFT) methods, such as LoRA and its variants, freeze pre-trained model weights \(W\) and inject learnable matrices \(ΔW\). These \(ΔW\) matrices are structured for efficient parameterization, often using techniques like low-rank approximations or scaling vectors.
Lingam, Vijay, Tejaswi, Atula, Vavre, Aditya, Shetty, Aneesh, Gudur, Gautham Krishna, Ghosh, Joydeep, Dimakis, Alex, Choi, Eunsol, Bojchevski, Aleksandar, Sanghavi, Sujay   +9 more
openaire   +2 more sources

Bistable Mechanisms 3D Printing for Mechanically Programmable Vibration Control

Advanced Engineering Materials, EarlyView.
This work introduces a 3D‐printed bistable mechanism integrated into tuned mass dampers (TMDs) for mechanically adaptive passive vibration suppression. Through optimized geometry, the bistable design provides adaptable vibration reduction across a broad range of scenarios, achieving effective vibration mitigation without complex controls or external ...
Ali Zolfagharian, Frédéric Demoly, Mohammad Lakhi, Bernard Rolfe, Mahdi Bodaghi   +4 more
wiley   +1 more source

Symbiotic Tuning: A Simple Approach for Enhancing Task Performance of Side-Tuning

IEEE Access
Reducing computational and memory overhead in fine-tuning large language models remains a significant challenge in natural language processing. While parameter-efficient fine-tuning (PEFT) methods, such as LoRA, have gained attention for reducing ...
Zhi-Quan Feng, Ying-Jia Lin, Kun-Ta Chuang, Hung-Yu Kao   +3 more
doaj   +1 more source

Parameter-Efficient Fine-Tuning for Foundation Models

25 pages, 6 figures, 7 ...
Zhang, Dan, Feng, Tao, Xue, Lilong, Wang, Yuandong, Dong, Yuxiao, Tang, Jie   +5 more
openaire   +2 more sources

Gradient Inversion Attacks on Parameter-Efficient Fine-Tuning

2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2025)
Sami, Hasin Us, Sen, Swapneel, Roy-Chowdhury, Amit K., Krishnamurthy, Srikanth V., Guler, Basak   +4 more
openaire   +2 more sources

Enhancing Corrosion Resistance and Mechanical Strength of 3D‐Printed Iron Polylactic Acid for Marine Applications via Laser Surface Texturing

Advanced Engineering Materials, EarlyView.
Laser surface texturing significantly improves the corrosion resistance and mechanical strength of 3D‐printed iron polylactic acid (Ir‐PLA) for marine applications. Optimal laser parameters reduce corrosion by 80% and enhance tensile strength by 25% and ductility by 15%.
Mohammad Rezayat, Mojtaba Karamimoghadam, Mohammadreza Lalegani Dezaki, Ali Zolfagharian, Giuseppe Casalino, Antonio Mateo, Mahdi Bodaghi   +6 more
wiley   +1 more source

Parameter-Efficient Continual Fine-Tuning: A Survey

The emergence of large pre-trained networks has revolutionized the AI field, unlocking new possibilities and achieving unprecedented performance. However, these models inherit a fundamental limitation from traditional Machine Learning approaches: their strong dependence on the \textit{i.i.d.} assumption hinders their adaptability to dynamic learning ...
Coleman, Eric Nuertey, Quarantiello, Luigi, Liu, Ziyue, Yang, Qinwen, Mukherjee, Samrat, Hurtado, Julio, Lomonaco, Vincenzo   +6 more
openaire   +2 more sources

Efficiency Gains in Broiler Production Through Contract Parameter Fine Tuning

Poultry Science, 1996
This paper reports on an analysis of existing broiler production contracts, with an attempt to establish the degree of efficiency gains possible from contract alteration. With the use of settlement cost and farm level data, an assessment is made of optimal grower input decisions given contract specifications.
T, Vukina, W E, Foster
openaire   +2 more sources

Trace Nickel Activated Biphasic Core‐CuOii/Shell‐CuOi Secondary Microspheres Enable Room Temperature Parts‐Per‐Trillion‐Level NO2 Detection

Advanced Engineering Materials, EarlyView.
An idea of designing novel sensors is proposed by creating appropriate Schottky barriers and vacancies between isomorphous Core‐CuOii/ Shell‐CuOi secondary microspheres and enhancing catalytic and spill‐over effects, and electronegativity via spontaneous biphasic separation, self‐assembly, and trace‐Ni‐doping.
Bala Ismail Adamu, Peipei Chen, Shafia Mukhtar Ibrahim, Mukhtar Lawan Adam, Attia Falak, Mokhotjwa Simon Dhlamini, Hanfu Wang, Lingxiang Jiang, Weiguo Chu   +8 more
wiley   +1 more source

Quantum-PEFT: Ultra parameter-efficient fine-tuning

ICLR ...
Koike-Akino, Toshiaki, Tonin, Francesco, Wu, Yongtao, Wu, Frank Zhengqing, Candogan, Leyla Naz, Cevher, Volkan   +5 more
openaire   +2 more sources