相关论文: Detailed Design and Optimization of Ferro-Electric…

We present a novel design of a FerroElectric Fast Reactive Tuner (FE-FRT) capable of modulating Mega VAR reactive power on a sub-microsecond time scale. We show detailed analytical estimates of the performance of this device and benchmark…

This work presents a state-of-the-art design of a Ferroelectric Fast Reactive Tuner (FE-FRT), capable of modulating high reactive power in TESLA type cavities on a microsecond time scale. The Mainz Energy-Recovering Superconducting…

This paper introduces a newly designed cavity tuner for superconducting radio-frequency (SRF) cavity. Aiming to overcome the drawbacks of traditional tuning systems, like the limited tuning range of piezoelectric tuner and the low-speed…

This paper describes performance of a voltage controlled phase shifter designed for a high average power and a high figure of merit. The device is a reflection-type, resonant phase shifter that utilizes ferroelectric capacitors and…

Recent advancements have illuminated the efficacy of some tensorization-decomposition Parameter-Efficient Fine-Tuning methods like LoRA and FacT in the context of Vision Transformers (ViT). However, these methods grapple with the challenges…

The Mixture-of-Experts (MoE) paradigm has emerged as a powerful approach for scaling transformers with improved resource utilization. However, efficiently fine-tuning MoE models remains largely underexplored. Inspired by recent works on…

Multimodal instruction tuning has proven to be an effective strategy for achieving zero-shot generalization by fine-tuning pre-trained Large Multimodal Models (LMMs) with instruction-following data. However, as the scale of LMMs continues…

Parameter-efficient tuning (PET) aims to transfer pre-trained foundation models to downstream tasks by learning a small number of parameters. Compared to traditional fine-tuning, which updates the entire model, PET significantly reduces…

Transformer-based large pre-trained models have shown remarkable generalization ability, and various parameter-efficient fine-tuning (PEFT) methods have been proposed to customize these models on downstream tasks with minimal computational…

This paper proposes a multivariable extremum seeking scheme using Fast Fourier Transform (FFT) for a network of subsystems working towards optimizing the sum of their local objectives, where the overall objective is the only available…

Field Programmable Gate Array (FPGA) is widely used in acceleration of deep learning applications because of its reconfigurability, flexibility, and fast time-to-market. However, conventional FPGA suffers from the tradeoff between chip area…

Tunable lasers are essential for optical communication, spectroscopy, and precision sensing, where flexible and fast control of the laser wavelength is needed. However, conventional tunable laser systems often rely on mechanical actuation,…

In the context of engineered quantum systems, there is a demand for superconducting tunable devices able to operate with high Q-factors at power levels equivalent to only a few photons. In this work, we developed a 3D microwave reentrant…

Machine learning-based compact models provide a rapid and efficient approach for estimating device behavior across multiple input parameter variations. In this study, we introduce two reverse-design algorithms that utilize these compact…

We demonstrate the active tuning of all-dielectric metasurfaces exhibiting high-quality factor (high-Q) resonances. The active control is provided by embedding the asymmetric silicon meta-atoms with liquid crystals, which allows the…

Multi-modal models excel in cross-modal tasks but are computationally expensive due to their billions of parameters. Parameter-efficient fine-tuning (PEFT) offers a solution by adding small trainable components while freezing pre-trained…

A fast and highly efficient frequency modulation at a high power level is described. The system incorporates ferroelectric phase shifters and a magic-T or a circulator. A magnetron may be considered as a potential application. The magnetron…

Adapting models pre-trained on large-scale datasets is a proven way to reach strong performance quickly for down-stream tasks. However, the growth of state-of-the-art mod-els makes traditional full fine-tuning unsuitable and difficult,…

The large models, as predicted by scaling raw forecasts, have made groundbreaking progress in many fields, particularly in natural language generation tasks, where they have approached or even surpassed human levels. However, the…

We present an integrated scheme for dielectric drive and read-out of high-Q nanomechanical resonators which enables tuning of both the resonance frequency and quality factor with an applied DC voltage. A simple model for altering these…