Dirk Hovy

Wisdom of Instruction-Tuned Language Model Crowds. Exploring Model Label Variation

2023 · arXiv (Cornell University)

Large Language Models (LLMs) exhibit remarkable text classification capabilities, excelling in zero- and few-shot learning (ZSL and FSL) scenarios. However, since they are trained on different datasets, performance varies widely across tasks between those models. …

Data-Efficient Strategies for Expanding Hate Speech Detection into Under-Resourced Languages

2022

Hate speech is a global phenomenon, but most hate speech datasets so far focus on English-language content. This hinders the development of more effective hate speech detection models in hundreds of languages spoken by billions …

Explaining Speech Classification Models via Word-Level Audio Segments and Paralinguistic Features

2023 · arXiv (Cornell University)

Recent advances in eXplainable AI (XAI) have provided new insights into how models for vision, language, and tabular data operate. However, few approaches exist for understanding speech models. Existing work focuses on a few spoken …

Comparing Pre-trained Human Language Models: Is it Better with Human Context as Groups, Individual Traits, or Both?

2024

Pre-trained language models consider the context of neighboring words and documents but lack any author context of the human generating the text.However, language depends on the author's states, traits, social, situational, and environmental attributes, collectively …

Demographic Factors Improve Classification Performance

2015

Extra-linguistic factors influence language use, and are accounted for by speakers and listeners.Most natural language processing (NLP) tasks to date, however, treat language as uniform.This assumption can harm performance.We investigate the effect of including demographic …