Trusting Your Evidence: Hallucinate Less with Context-aware Decoding
Weijia Shi, Xiaochuang Han, Mike Lewis, Yulia Tsvetkov, Luke Zettlemoyer, Wen-tau Yih
Abstract
Language models (LMs) often struggle to pay enough attention to the input context, and generate texts that are unfaithful or contain hallucinations. To mitigate this issue, we present context-aware decoding (CAD), which follows a contrastive output distribution that amplifies the difference between the output probabilities when a model is used with and without context. Our experiments show that CAD, without additional training, significantly improves the faithfulness of different LM families, including OPT, GPT, LLaMA, and FLAN-T5 for summarization tasks (e.g., 14.3% gain for LLaMA in factuality metrics). Furthermore, CAD is particularly effective in overriding a model’s prior knowledge when it contradicts the provided context, leading to substantial improvements in tasks where resolving the knowledge conflict is essential. Our code is publicly released at https://github.com/xhan77/context-aware-decoding.
BibTeX
@inproceedings{shi-etal-2024-trusting,
title = "Trusting Your Evidence: Hallucinate Less with Context-aware Decoding",
author = "Shi, Weijia and
Han, Xiaochuang and
Lewis, Mike and
Tsvetkov, Yulia and
Zettlemoyer, Luke and
Yih, Wen-tau",
editor = "Duh, Kevin and
Gomez, Helena and
Bethard, Steven",
booktitle = "Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)",
month = jun,
year = "2024",
address = "Mexico City, Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.naacl-short.69/",
doi = "10.18653/v1/2024.naacl-short.69",
pages = "783--791"
}