# Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information

arXiv: https://arxiv.org/abs/2404.09480 · HTML: https://mlsanity.com/paper/2404.09480/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 24 references that are in the map.
- First posted on arXiv 15 Apr 2024.

## Builds on (most cited first, 16 of 16)

- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization](https://mlsanity.com/paper/1808.08745/) (2018)
- [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://mlsanity.com/paper/1912.08777/) (2019)
- [Accelerating Large Language Model Decoding with Speculative Sampling](https://mlsanity.com/paper/2302.01318/) (2023)
- [BARTScore: Evaluating Generated Text as Text Generation](https://mlsanity.com/paper/2106.11520/) (2021)
- [Evaluating the Factual Consistency of Abstractive Text Summarization](https://mlsanity.com/paper/1910.12840/) (2019)
- [How Language Model Hallucinations Can Snowball](https://mlsanity.com/paper/2305.13534/) (2023)
- [AlignScore: Evaluating Factual Consistency with a Unified Alignment Function](https://mlsanity.com/paper/2305.16739/) (2023)
- [Surface Form Competition: Why the Highest Probability Answer Isn't Always Right](https://mlsanity.com/paper/2104.08315/) (2021)
- [On Hallucination and Predictive Uncertainty in Conditional Language Generation](https://mlsanity.com/paper/2103.15025/) (2021)
- [Detecting Hallucinated Content in Conditional Neural Sequence Generation](https://mlsanity.com/paper/2011.02593/) (2020)
- [Mutual Information Alleviates Hallucinations in Abstractive Summarization](https://mlsanity.com/paper/2210.13210/) (2022)
- [FactPEGASUS: Factuality-Aware Pre-training and Fine-tuning for Abstractive Summarization](https://mlsanity.com/paper/2205.07830/) (2022)
- [Constrained Abstractive Summarization: Preserving Factual Consistency with Constrained Generation](https://mlsanity.com/paper/2010.12723/) (2020)
- [Don't Say What You Don't Know: Improving the Consistency of Abstractive Summarization by Constraining Beam Search](https://mlsanity.com/paper/2203.08436/) (2022)

## Built on by (most cited first, 6 of 6)

- [Model-based Preference Optimization in Abstractive Summarization without Human Feedback](https://mlsanity.com/paper/2409.18618/) (2024)
- [What Makes a Good Natural Language Prompt?](https://mlsanity.com/paper/2506.06950/) (2025)
- [TIAR: Trajectory-Informed Advantage Reweighting for LLM Abstention Learning](https://mlsanity.com/paper/2605.25850/) (2026)
- [NTS-CoT: Mitigating Hallucinations in LLM-based News Timeline Summarization with Chain-of-Thought Reasoning](https://mlsanity.com/paper/2606.13171/) (2026)
- [Decoupling Generation and Selection for Budget-Constrained Faithful Summarization](https://mlsanity.com/paper/2608.03655/) (2026)
- [Does This Action Still Explain the Task? Reverse Scoring for Diffusion Language Model Agents](https://mlsanity.com/paper/2609.38536/) (2026)

## Abstract (arXiv)

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, neglecting the details of the source text. To alleviate this model bias, we introduce a decoding strategy based on domain-conditional pointwise mutual information. This strategy adjusts the generation probability of each token by comparing it with the token's marginal probability within the domain of the source text. According to evaluation on the XSUM dataset, our method demonstrates improvement in terms of faithfulness and source relevance. The code is publicly available at urlhttps://github.com/qqplot/dcpmi.