Mutual information alleviates hallucinations in abstractive summarization
L Van der Poel, R Cotterell, C Meister - arXiv preprint arXiv:2210.13210, 2022 - arxiv.org
Despite significant progress in the quality of language generated from abstractive
summarization models, these models still exhibit the tendency to hallucinate, ie, output …
summarization models, these models still exhibit the tendency to hallucinate, ie, output …
Don't say what you don't know: Improving the consistency of abstractive summarization by constraining beam search
Abstractive summarization systems today produce fluent and relevant output, but often"
hallucinate" statements not supported by the source text. We analyze the connection …
hallucinate" statements not supported by the source text. We analyze the connection …
Learning with rejection for abstractive text summarization
State-of-the-art abstractive summarization systems frequently hallucinate content that is not
supported by the source document, mainly due to noise in the training dataset. Existing …
supported by the source document, mainly due to noise in the training dataset. Existing …
Training dynamics for text summarization models
Pre-trained language models (eg BART) have shown impressive results when fine-tuned on
large summarization datasets. However, little is understood about this fine-tuning process …
large summarization datasets. However, little is understood about this fine-tuning process …
Constrained abstractive summarization: Preserving factual consistency with constrained generation
Despite significant progress, state-of-the-art abstractive summarization methods are still
prone to hallucinate content inconsistent with the source document. In this paper, we …
prone to hallucinate content inconsistent with the source document. In this paper, we …
Faithfulness-aware decoding strategies for abstractive summarization
Despite significant progress in understanding and improving faithfulness in abstractive
summarization, the question of how decoding strategies affect faithfulness is less studied …
summarization, the question of how decoding strategies affect faithfulness is less studied …
Alleviating exposure bias via contrastive learning for abstractive text summarization
Encoder-decoder models have achieved remarkable success in abstractive text
summarization, which aims to compress one or more documents into a shorter version …
summarization, which aims to compress one or more documents into a shorter version …
Understanding neural abstractive summarization models via uncertainty
An advantage of seq2seq abstractive summarization models is that they generate text in a
free-form manner, but this flexibility makes it difficult to interpret model behavior. In this work …
free-form manner, but this flexibility makes it difficult to interpret model behavior. In this work …
Diversity driven attention model for query-based abstractive summarization
Abstractive summarization aims to generate a shorter version of the document covering all
the salient points in a compact and coherent fashion. On the other hand, query-based …
the salient points in a compact and coherent fashion. On the other hand, query-based …
Length control in abstractive summarization by pretraining information selection
Previous length-controllable summarization models mostly control lengths at the decoding
stage, whereas the encoding or the selection of information from the source document is not …
stage, whereas the encoding or the selection of information from the source document is not …