Measuring the distance between machine-produced and human language is a critical open problem. Inspired by empirical findings from psycholinguistics on the periodicity of entropy in language, we propose FACE, a set of metrics based on Fourier Analysis of the estimated Cross-Entropy of language, for measuring the similarity between model-generated and human-written languages. Based on an open-ended generation task and the experimental data from previous studies, we find that FACE can effectively identify the human-model gap, scales with model size, reflects the outcomes of different sampling methods for decoding, correlates well with other evaluation metrics and with human judgment scores.
@article{arxiv.2305.10307,
title = {FACE: Evaluating Natural Language Generation with Fourier Analysis of Cross-Entropy},
author = {Zuhao Yang and Yingfang Yuan and Yang Xu and Shuo Zhan and Huajun Bai and Kefan Chen},
journal= {arXiv preprint arXiv:2305.10307},
year = {2023}
}