Faithfulness-Aware Decoding Strategies for Abstractive Summarization

Abstract

Despite significant progress in understanding and improving faithfulness in abstractive summarization, the question of how decoding strategies affect faithfulness is less studied. We present a systematic study of the effect of generation techniques such as beam search and nucleus sampling on faithfulness in abstractive summarization. We find a consistent trend where beam search with large beam sizes produces the most faithful summaries while nucleus sampling generates the least faithful ones. We propose two faithfulness-aware generation methods to further improve faithfulness over current generation techniques: (1) ranking candidates generated by beam search using automatic faithfulness metrics and (2) incorporating lookahead heuristics that produce a faithfulness score on the future summary. We show that both generation methods significantly improve faithfulness across two datasets as evaluated by four automatic faithfulness metrics and human evaluation. To reduce computational cost, we demonstrate a simple distillation approach that allows the model to generate faithful summaries with just greedy decoding. 1 * * Work conducted during an internship at Amazon.

Faithful AI in Medicine: A Systematic Review with Large Language Models and Beyond

“…Medical Knowledge Medical knowledge is critical for alleviating the hallucination of medical AI methods for various tasks. Future efforts should explore effective ways of injecting medical knowledge into medical AI methods, such as encoding medical knowledge based on prompt learning 106 , and explicitly leveraging the medical knowledge to guide the text generation 87 .…”

“…However, the medical knowledge are typically incorporated in implicit ways based on RL, we consider that future efforts should pay more attention to explicitly incorporating medical knowledge on improving the encoder and decoder of the LLMs. For example, investigating the use knowledge grounded backbone language models as encoder 37 , and developing decoding strategy guided by medical facts 55,87 . Moreover, the radiology report generation task requires the combination of information both radiology images and the associated text reports, we believe cross-modality vision-language foundation models 88 should be explored to improve the faithfulness of radiology report generation methods in the future.…”

The Curious Case of Hallucinatory (Un)answerability: Finding Truths in the Hidden States of Over-Confident Large Language Models

“…Furthermore, recent studies have suggested utilizing recent advances in white-box model interpretability (Geva et al, 2022b;Li et al, 2022;Mallen et al, 2022;Mickus et al, 2022;Meng et al, 2023;Geva et al, 2023) and probing (Adi et al, 2017;Conneau et al, 2018;Voita et al, 2019;Slobodkin et al, 2021) for manipulating the model predictions and analyzing when LLMs struggle to answer questions. Recent works also tried to use beam search decoding to manipulate the generated outputs by using the information encapsulated in several beams (Meister et al, 2020;Leblond et al, 2021;Slobodkin et al, 2023;Wan et al, 2023b). Finally, early exiting in language models (Schwartz et al, 2020;Schuster et al, 2022;Din et al, 2023) and model prediction calibration (Desai and Durrett, 2020;Jiang et al, 2021;Dhuliawala et al, 2022;Geva et al, 2022a) are strongly related to our work, as they suggest to analyze and improve the model predictions and output distribution.…”