DecipherPref: Analyzing Influential Factors in Human Preference Judgments via GPT-4
Submission Type: Regular Long Paper
Submission Track: Summarization
Keywords: Large Language Models, Summarization Evaluation, Human Preference Judgments, GPT-4
TL;DR: A comprehensive analysis of a dataset of pairwise human judgments released by OpenAI, using the Bradley-Terry-Luce (BTL) model to uncover key factors that may influence human preferences.
Abstract: Human preference judgments are pivotal in guiding large language models (LLMs) to produce outputs that align with human values. Human evaluations are also used in summarization tasks to compare outputs from various systems, complementing existing automatic metrics. Despite their significance, however, there has been limited research probing these pairwise or $k$-wise comparisons. The collective impact and relative importance of factors such as output length, informativeness, fluency, and factual consistency are still not well understood. It is also unclear if there are other hidden factors influencing human judgments. In this paper, we conduct an in-depth examination of a collection of pairwise human judgments released by OpenAI. Utilizing the Bradley-Terry-Luce (BTL) model, we reveal the inherent preferences embedded in these human judgments. We find that the most favored factors vary across tasks and genres, whereas the least favored factors tend to be consistent, e.g., outputs are too brief, contain excessive off-focus content or hallucinated facts. Our findings have implications on the construction of balanced datasets in human preference evaluations, which is a crucial step in shaping the behaviors of future LLMs.
Submission Number: 124
Loading