CaPE: Contrastive Parameter Ensembling for Reducing Hallucination in Abstractive Summarization

Prafulla Kumar Choubey, Alex Fabbri, Jesse Vig, Chien-Sheng Wu, Wenhao Liu, Nazneen Rajani


Abstract
Hallucination is a known issue for neural abstractive summarization models. Recent work suggests that the degree of hallucination may depend on factual errors in the training data. In this work, we propose a new method called Contrastive Parameter Ensembling (CaPE) to use training data more effectively, utilizing variations in noise in training samples to reduce hallucination. Starting with a base model fine-tuned on an entire dataset, we additionally train expert and anti-expert models on clean and noisy subsets of the data, respectively. We then adjust the parameters of the base model by adding (subtracting) the parameters of the expert (anti-expert), advancing the recent work on additive parameter ensembling approaches. Trained on a much smaller data subset, expert and anti-expert models only fractionally (<14%) increases the total training time. Further, CaPE uses parameter ensembling and does not increase the inference time. Experimental results show that CaPE improves performance across different automatic factual metrics and human evaluation, with a maximum improvement of 16.69% and 15.38% on summary-level dependency-arc entailment accuracy for the XSUM and CNN/DM datasets. The CaPE model performs comparably to the base model on metrics of informativeness such as ROUGE.
Anthology ID:
2023.findings-acl.685
Volume:
Findings of the Association for Computational Linguistics: ACL 2023
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
10755–10773
Language:
URL:
https://aclanthology.org/2023.findings-acl.685
DOI:
10.18653/v1/2023.findings-acl.685
Bibkey:
Cite (ACL):
Prafulla Kumar Choubey, Alex Fabbri, Jesse Vig, Chien-Sheng Wu, Wenhao Liu, and Nazneen Rajani. 2023. CaPE: Contrastive Parameter Ensembling for Reducing Hallucination in Abstractive Summarization. In Findings of the Association for Computational Linguistics: ACL 2023, pages 10755–10773, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
CaPE: Contrastive Parameter Ensembling for Reducing Hallucination in Abstractive Summarization (Choubey et al., Findings 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-3/2023.findings-acl.685.pdf