Results 91 to 100 of about 7,279,238 (253)

Combining Neural Architecture Search and Automatic Code Optimization: A Survey [PDF]

Deep Learning models have experienced exponential growth in complexity and resource demands in recent years. Accelerating these models for efficient execution on resource-constrained devices has become more crucial than ever.
Ouarnoughi, Hamza, Baghdadi, Riyadh, Bachiri, Inas, Niar, Smail, Benmeziane, Hadjer, Aries, Abdelkrime   +5 more
core  

Precision Nanomedicine for Traumatic Neural Injury: Functional Paradigms and Translational Frameworks

Advanced Science, EarlyView.
This review presents a pathology‐driven framework unifying nanomedicine across TBI, SCI, and PNI. Epidemiological evidence and comparative single‐cell transcriptomic reanalysis define shared pathological modules, injury‐specific effectors, and therapeutic bottlenecks. Six functional paradigms connect nanomaterial structure and biological fate to injury‐
Mouyuan Sun, Shuangyang Li, Zhixu He, Jingyi Xie, Yaxian Luo, Xuankai Fan, Jingyu Zhang, Luying Qin, Lianjie Peng, Tao Qiu, Tian Zhang, Huiming Wang, Mengfei Yu   +12 more
wiley   +1 more source

LLM-NAS: LLM-driven Hardware-Aware Neural Architecture Search

Proceedings of the Genetic and Evolutionary Computation Conference Companion
Hardware-Aware Neural Architecture Search (HW-NAS) requires joint optimization of accuracy and latency under device constraints. Traditional supernet-based methods require multiple GPU days per dataset. Large Language Model (LLM)-driven approaches avoid training a large supernet and can provide quick feedback, but we observe an exploration bias: the ...
Hengyi Zhu, Grace Li Zhang, Shaoyi Huang
openaire   +2 more sources

Metasurface‐Based Multidimensional Light‐Field Manipulation: From Technologies to Applications

Advanced Science, EarlyView.
Multidimensional metasurfaces enable versatile manipulation of multiple optical degrees of freedom, including phase, amplitude, polarization, wavelength, and orbital angular momentum. This review summarizes their fundamental mechanisms, emerging strategies for multidimensional light‐field control, and key applications in optical computing, imaging ...
Heliang Yin, Shuo Sun, Jiahao Zhang, Jianuo Hu, Yubin Qiao, Liang Chen, Jin Li   +6 more
wiley   +1 more source

MicroNAS: Zero-Shot Neural Architecture Search for MCUs [PDF]

Neural Architecture Search (NAS) effectively discovers new Convolutional Neural Network (CNN) architectures, particularly for accuracy optimization.
Qiao, Ye, Zhang, Yifan, Xu, Haocheng, Huang, Sitao   +3 more
core   +1 more source

Highly Linear Perovskite Synapses for Neuromorphic Computing

Advanced Science, EarlyView.
Voltage spikes applied to an Au/ACI‐phase perovskite/ITO memristor emulate the ionic signaling of a biological synapse, generating a tunable postsynaptic current. The vertically aligned perovskite film confines iodide‐vacancy migration, enabling highly linear, and symmetric conductance updates with 100 states and 5.5 aJ per event. Such synapses sustain
Zhongwen Sun, Boyu Wan, Zifan Zhang, Zhikai Wang, Xuan Zhao, Haonan Si, Qingliang Liao, Yue Zhang   +7 more
wiley   +1 more source

Neutron Depth Profiling: The Core Link Connecting the Microscopic Mechanisms of Lithium‐Ion Batteries With Intelligent Design

Advanced Science, EarlyView.
Neutron depth profiling exploits the 6Li(n,3H)4He reaction to provide quantitative, isotope‐specific, depth‐resolved maps of lithium inside operating lithium‐ion batteries, directly linking lithium transport, dendrite growth and interfacial short‐circuiting to a physics‐informed machine‐learning roadmap that turns this non‐destructive operando ...
Yulong Zhang, Chunjing Yue, Shijiao Liu, Muyang Chen, Xiaokang Shi, Rui Tan, Hongqing Hao, Jie Yang, Xinhua Liu   +8 more
wiley   +1 more source

A Dual‐Branch Flux‐Based Extended Memristor Model With Machine‐Learning‐Assisted Calibration

Advanced Electronic Materials, EarlyView.
Multilayer oxide memristors integrated in crossbar arrays are described through a dual‐branch, flux‐controlled compact model. A three‐stage calibration workflow combining Latin hypercube sampling, Bayesian optimization, and gradient‐based refinement extracts device parameters from experimental data.
Davide Rossetti, K. Nikiruy, A.S. Demirkol, Alon Ascoli, M. Ziegler, R. Tetzlaff, Fernando Corinto   +6 more
wiley   +1 more source

Blood‐Based Biomarkers in Alzheimer's Disease: Advances in Detection Technologies and Clinical Applications

AGING MEDICINE, EarlyView.
Blood‐based biomarkers are emerging as accessible tools for Alzheimer's disease screening, diagnosis, risk stratification, and longitudinal monitoring. This review summarizes major biomarkers, including Aβ42/Aβ40, phosphorylated tau species, NfL, GFAP, genetic risk markers, and exploratory inflammatory, synaptic, and proteomic candidates.
Shuangshuang Li, Lin Huang, Qihao Guo
wiley   +1 more source

Evolution of Physical Intelligence Across Scales

Advanced Intelligent Discovery, EarlyView.
By following the evolution of physical intelligence across scales, this article shows how intelligence arises from materials, structures, physical interactions, and collectives. It establishes physical intelligence as the evolutionary foundation upon which embodied intelligence is built.
Ke Liu, Tianyun Huang, Aming Li, Pengyu Lv, Tingting Qin, Wenhui Chen, Hui Wei, Huiling Duan   +7 more
wiley   +1 more source