Benchmarking Large Language Models for News Summarization
Published
2024-02-03
Tianyi Zhang
,
Faisal Ladhak
,
Esin Durmus
,
Percy Liang
,
Kathleen McKeown
,
Tatsunori Hashimoto
Tianyi Zhang
Stanford University
Faisal Ladhak
Columbia University
Esin Durmus
Anthropic, Stanford University
Percy Liang
Stanford University
Kathleen McKeown
Columbia University
Tatsunori Hashimoto
Stanford University
Abstract
Large language models (LLMs) have shown promise for automatic summarization but the reasons behind their successes are poorly understood. By conducting a human evaluation on ten LLMs across different pretraining methods, prompts, and model scales, we make two important observations. First, we find instruction tuning, and not model size, is the key to the LLM's zero-shot summarization capability. Second, existing studies have been limited by low-quality references, leading to underestimates of human performance and lower few-shot and finetuning performance. To better evaluate LLMs, we perform human evaluation over high-quality summaries we collect from freelance writers. Despite major stylistic differences such as the amount of paraphrasing, we find that LMM summaries are judged to be on par with human written summaries.
Article at MIT Press
Presented at EMNLP 2023