We propose a parameter sharing method for Transformers (Vaswani et al., 2017). The
proposed approach relaxes a widely used technique, which shares parameters for one layer …

We often use perturbations to regularize neural models. For neural encoder-decoders,
previous studies applied the scheduled sampling (Bengio et al., 2015) and adversarial …

We propose a novel data-augmentation technique for neural machine translation based on
ROT-$ k $ ciphertexts. ROT-$ k $ is a simple letter substitution cipher that replaces a letter in …

Graph neural networks (GNNs) are deep learning models designed specifically for graph
data, and they typically rely on node features as the input to the first layer. When applying …

Understanding the influence of a training instance on a neural network model leads to
improving interpretability. However, it is difficult and inefficient to evaluate the influence …

This paper empirically investigates the relationship between subword vocabulary size and
the performance of large language models (LLMs) to provide insights on how to define the …

There is a growing interest in utilizing large-scale language models (LLMs) to advance next-
generation Recommender Systems (RecSys), driven by their outstanding language …

A modular approach has the advantage of being compositional and controllable, comparing
to most end-to-end models. In this paper we propose Extract-Select-Rewrite (ESR), a three …

Automatic text summarization is a lucrative field in natural language processing (NLP). The
amount of data flow has multiplied with the switch to digital. The massive datasets hold a …

The field of Natural Language Processing (NLP) has advanced significantly in recent years,
resulting in the creation of complex systems that indicate a deep comprehension of human …