Abstract: Described herein are nanoribbon-based transistors with a highly uniform oxide layer around semiconductor nanoribbon channels, and a high-pressure steam process for growth the oxide layer. The high-pressure steam process is a self-limiting process that results in a more uniform oxide than standard deposition or implantation methods. The uniformity enables greater control over oxide thickness, with improved breakdown voltages and drive currents.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for performing a machine learning task on an input sequence of characters that has a respective character at each of a plurality of character positions to generate a network output. One of the systems includes a neural network configured to perform the machine learning task, the neural network comprising a gradient-based sub-word tokenizer and an output neural network. The gradient-based sub-word tokenizer is configured to apply a learned, i.e., flexible, sub-word tokenization strategy to the input sequence of characters to generate a sequence of latent sub-word representations. The output neural network is configured to process the latent sub-word representation to generate the network output for the task.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating output sequences using auto-regressive decoder neural networks. In particular, during generation, adaptive early exiting is used to reduce the time required to generate the output sequence.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating output sequences using auto-regressive decoder neural networks. In particular, during generation, adaptive early exiting is used to reduce the time required to generate the output sequence.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating output sequences using auto-regressive decoder neural networks. In particular, during generation, adaptive early exiting is used to reduce the time required to generate the output sequence.

Abstract: A microelectronic assembly and a method of forming same. The assembly includes: first and second microelectronic structures; and an interface layer between the two microelectronic structures including dielectric portions in registration with dielectric layers of each of the microelectronic structures, and electrically conductive portions in registration with electrically conductive structures of each of the microelectronic structures, wherein the dielectric portions include an oxide of a metal, and the electrically conductive portions include the metal.

Abstract: Provided are machine-learned attention models that feature omnidirectional processing, example implementations of which can be referred to as Omnidirectional Representations from Transformers (OMNINET). In example models described in the present disclosure, instead of maintaining a strictly horizontal receptive field, each token is allowed to attend to all tokens in some or all of the other tokens across the entire network.

Abstract: Disclosed herein are stacked transistors with dielectric between source/drain materials of different strata, as well as related methods and devices. In some embodiments, an integrated circuit structure may include stacked strata of transistors, wherein a dielectric material is between source/drain materials of adjacent strata, and the dielectric material is conformal on underlying source/drain material.