1

Decoder-hybrid-decoder architecture for efficient reasoning with long generation

Streaming Sequence Transduction through Dynamic Compression

Upsample or Upweight? Balanced Training on Heavily Imbalanced Datasets

X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scale

Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation

Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?

Narrowing the Gap between Zero- and Few-shot Machine Translation by Matching Styles

JHU IWSLT 2024 Dialectal and Low-resource System Description

The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts

A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models