We propose the In-context Autoencoder (ICAE) for context compression in ...
Human intelligence thrives on the concept of cognitive synergy, where
co...
Enhancing word usage is a desired feature for writing assistance. To fur...
Effectively utilizing LLMs for complex tasks is challenging, often invol...
Semiparametric language models (LMs) have shown promise in continuously
...
Numerous dual-energy CT (DECT) techniques have been developed in the pas...
We introduce PoliteRewrite – a dataset for polite language rewrite
which...
We propose eXtensible Prompt (X-Prompt) for prompting a large language m...
Data-driven models, such as FourCastNet (FCN), have shown exemplary
perf...
We study lossless acceleration for seq2seq generation with a novel decod...
Text revision refers to a family of natural language generation tasks, w...
In this paper, we propose Generalized Aggressive Decoding (GAD) – a nove...
We propose EdgeFormer – a parameter-efficient Transformer of the
encoder...
Synthetic data construction of Grammatical Error Correction (GEC) for
no...
Recent studies on compression of pretrained language models (e.g., BERT)...
Dual-energy CT (DECT) has been widely investigated to generate more
info...
In this paper, we propose Shallow Aggressive Decoding (SAD) to improve t...
Cant is important for understanding advertising, comedies and dog-whistl...
In this paper, we generalize text infilling (e.g., masked language model...
We propose a novel language-independent approach to improve the efficien...
In this paper, we propose Patience-based Early Exit, a straightforward y...
The main barrier to progress in the task of Formality Style Transfer is ...
In this paper, we introduce DropHead, a structured dropout method
specif...
In this paper, we propose a novel model compression approach to effectiv...
Local sequence transduction (LST) tasks are sequence transduction tasks ...
Conventional Generative Adversarial Networks (GANs) for text generation ...
Sentence Split and Rephrase aims to break down a complex sentence into
s...
Non-smooth regularization is widely used in image reconstruction to elim...
We propose a novel data synthesis method to generate diverse error-corre...
We study sequence-to-sequence (seq2seq) pre-training with data augmentat...
Formality style transformation is the task of modifying the formality of...
Neural sequence-to-sequence (seq2seq) approaches have proven to be succe...
In the medical domain, identifying and expanding abbreviations in clinic...