Related papers: Integrated electro-optic attention nonlinearities …
The quadratic computational complexity of softmax transformers has become a bottleneck in long-context scenarios. In contrast, linear attention model families provide a promising direction towards a more efficient sequential model. These…
In recent years, Transformer networks are beginning to replace pure convolutional neural networks (CNNs) in the field of computer vision due to their global receptive field and adaptability to input. However, the quadratic computational…
In this work we provide new insights into the transformer architecture, and in particular, its best-known variant, BERT. First, we propose a method to measure the degree of non-linearity of different elements of transformers. Next, we focus…
Transformer-based large language models (LLMs) exhibit impressive performance in generative tasks but also introduce significant challenges in real-world serving due to inefficient use of the expensive, computation-optimized accelerators.…
Transformers' quadratic computational complexity limits their scalability despite remarkable performance. While linear attention reduces this to linear complexity, pre-training such models from scratch remains, in most cases, prohibitively…
Acousto-optic (AO) interactions provide a powerful interface between the microwave and optical domains, enabling functionalities such as optical switching, non-reciprocal propagation and efficient microwave-to-optical transduction.…
Vision Transformer(ViT) is now dominating many vision tasks. The drawback of quadratic complexity of its token-wise multi-head self-attention (MHSA), is extensively addressed via either token sparsification or dimension reduction (in…
From the onset of high-speed optical communications, lithium niobite (LN) has been the material of choice for electro-optic modulators owing to its large electro-optic response, wide transparent window, excellent thermal stability and…
Transformers have revolutionized deep learning in numerous fields, including natural language processing, computer vision, and audio processing. Their strength lies in their attention mechanism, which allows for the discovering of complex…
The rapid expansion of cloud computing and artificial intelligence has driven the demand for faster optical components in data centres to unprecedented levels. A key advancement in this field is the integration of multiple photonic…
Integrated electro-optic (EO) modulators are fundamental photonics components with utility in domains ranging from digital communications to quantum information processing. At telecommunication wavelengths, thin-film lithium niobate…
The quadratic cost of softmax attention limits Transformer scalability in high-resolution vision. We introduce Infinite Self-Attention (InfSA), a spectral reformulation that treats each attention layer as a diffusion step on a…
Ultrathin plasmonic films that approach the trans-dimensional (TD) thickness limit provide a promising route for light_matter interaction control and manipulation, yet their nonlinear optical response near the epsilon_near_zero (ENZ)…
In the past decade, photonic integrated circuits (PICs) based on thin-film lithium niobate (TFLN) have advanced in various fields, including optical communication, nonlinear photonics, and quantum optics. A critical component is an…
A low voltage, wide bandwidth compact electro-optic modulator is a key building block in the realization of tomorrow's communication and networking needs. Recent advances in the fabrication and application of thin-film lithium niobate, and…
Optomechanical transduction harnesses the interaction between optical fields and mechanical motion to achieve sensitive measurement of weak mechanical quantities with inherently low noise. Lithium niobate combines low optical loss, strong…
Deploying Vision Transformers (ViTs) on near-sensor analog accelerators demands training pipelines that are explicitly aligned with device-level noise and energy constraints. We introduce a compact framework for silicon-photonic execution…
Spatial light modulators (SLMs) are devices for modulating amplitude, phase or polarization of a light beam on demand. Such devices have been playing an indispensable inuence in many areas from our daily entertainments to scientific…
Modern communication and sensing technologies rely on connecting the optical domain with the microwave domain. Terahertz technologies, spanning increased frequencies from 100 GHz to 10 THz, are critical for providing larger bandwidths and…
Thanks to its high quadratic nonlinear susceptibilty and low propagation losses, thin film lithium niobate (TFLN) on insulator is an ideal platform for laser frequency conversion and generation of quantum states of light. Frequency…