Jaepill Choi — accepted AI-conference papers

2024

Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information

Kyubyung Chae, Jaepill Choi, Yohan Jo, Taesup Kim

NAACL 2024findings

A primary challenge in abstractive summarization is hallucination—the phenomenon where a model generates plausible text that is absent in the source text. We hypothesize that the domain (or topic) of the source text triggers the model to generate text that is highly probable in the domain, neglectin…

Cited by 3 Source ↗PDF ↗code ↗Scholar ↗

2024

Model-based Preference Optimization in Abstractive Summarization without Human Feedback

Jaepill Choi, Kyubyung Chae, Jiwoo Song, Yohan Jo, Taesup Kim

EMNLP 2024main

In abstractive summarization, the challenge of producing concise and accurate summaries arises from the vast amount of information contained in the source document. Consequently, although Large Language Models (LLMs) can generate fluent text, they often introduce inaccuracies by hallucinating conten…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗