Open-Domain Text Evaluation via Contrastive Distribution Methods
Abstract
Recent advancements in open-domain text generation, driven by the power of large pre-trained language models (LLMs), have demonstrated remarkable performance. However, assessing these models' generation quality remains a challenge. In this paper, we introduce a novel method for evaluating open-domain text generation called Contrastive Distribution Methods (CDM). Leveraging the connection between increasing model parameters and enhanced LLM performance, CDM creates a mapping from the _contrast_ of two probabilistic distributions -- one known to be superior to the other -- to quality measures. We investigate CDM for open-domain text generation evaluation under two paradigms: 1) _Generative_ CDM, which harnesses the contrast of two language models' distributions to generate synthetic examples for training discriminator-based metrics; 2) _Discriminative_ CDM, which directly uses distribution disparities between two language models for evaluation. Our experiments on coherence evaluation for multi-turn dialogue and commonsense evaluation for controllable generation demonstrate CDM's superior correlate with human judgment than existing automatic evaluation metrics, highlighting the strong performance and generalizability of our approach.
Cite
@article{arxiv.2306.11879,
title = {Open-Domain Text Evaluation via Contrastive Distribution Methods},
author = {Sidi Lu and Hongyi Liu and Asli Celikyilmaz and Tianlu Wang and Nanyun Peng},
journal= {arXiv preprint arXiv:2306.11879},
year = {2024}
}
Comments
Accepted to ICML 2024