# On Positional Bias of Faithfulness for Long-form Summarization

arXiv: https://arxiv.org/abs/2410.23609 · HTML: https://mlsanity.com/paper/2410.23609/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- Builds on 43 of its 52 references that are in the map.
- First posted on arXiv 31 Oct 2024.

## Builds on (most cited first, 30 of 43)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization](https://mlsanity.com/paper/1808.08745/) (2018)
- [Get To The Point: Summarization with Pointer-Generator Networks](https://mlsanity.com/paper/1704.04368/) (2017)
- [What Makes Good In-Context Examples for GPT-3?](https://mlsanity.com/paper/2101.06804/) (2021)
- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [Large Language Models are not Fair Evaluators](https://mlsanity.com/paper/2305.17926/) (2023)
- [SummEval: Re-evaluating Summarization Evaluation](https://mlsanity.com/paper/2007.12626/) (2020)
- [Large Language Models Are Not Robust Multiple Choice Selectors](https://mlsanity.com/paper/2309.03882/) (2023)
- [A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents](https://mlsanity.com/paper/1804.05685/) (2018)
- [Benchmarking Large Language Models for News Summarization](https://mlsanity.com/paper/2301.13848/) (2023)
- [Multi-News: a Large-Scale Multi-Document Summarization Dataset and Abstractive Hierarchical Model](https://mlsanity.com/paper/1906.01749/) (2019)
- [Is ChatGPT a Good NLG Evaluator? A Preliminary Study](https://mlsanity.com/paper/2303.04048/) (2023)
- [SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in Summarization](https://mlsanity.com/paper/2111.09525/) (2021)
- [QMSum: A New Benchmark for Query-based Multi-domain Meeting Summarization](https://mlsanity.com/paper/2104.05938/) (2021)
- [Efficient Attentions for Long Document Summarization](https://mlsanity.com/paper/2104.02112/) (2021)
- [News Summarization and Evaluation in the Era of GPT-3](https://mlsanity.com/paper/2209.12356/) (2022)
- [AlignScore: Evaluating Factual Consistency with a Unified Alignment Function](https://mlsanity.com/paper/2305.16739/) (2023)
- [Long-context LLMs Struggle with Long In-context Learning](https://mlsanity.com/paper/2404.02060/) (2024)
- [MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents](https://mlsanity.com/paper/2404.10774/) (2024)
- [QAFactEval: Improved QA-Based Factual Consistency Evaluation for Summarization](https://mlsanity.com/paper/2112.08542/) (2021)
- [Revisiting the Gold Standard: Grounding Summarization Evaluation with Robust Human Evaluation](https://mlsanity.com/paper/2212.07981/) (2022)
- [BooookScore: A systematic exploration of book-length summarization in the era of LLMs](https://mlsanity.com/paper/2310.00785/) (2023)
- [When Benchmarks are Targets: Revealing the Sensitivity of Large Language Model Leaderboards](https://mlsanity.com/paper/2402.01781/) (2024)
- [Hallucinated but Factual! Inspecting the Factuality of Hallucinations in Abstractive Summarization](https://mlsanity.com/paper/2109.09784/) (2021)
- [Understanding Factual Errors in Summarization: Errors, Summarizers, Datasets, Error Detectors](https://mlsanity.com/paper/2205.12854/) (2022)
- [Multi-XScience: A Large-scale Dataset for Extreme Multi-document Summarization of Scientific Articles](https://mlsanity.com/paper/2010.14235/) (2020)
- [LongEval: Guidelines for Human Evaluation of Faithfulness in Long-form Summarization](https://mlsanity.com/paper/2301.13298/) (2023)
- [Do Long-Range Language Models Actually Use Long-Range Context?](https://mlsanity.com/paper/2109.09115/) (2021)

## Built on by (most cited first, 22 of 22)

- [Does Context Matter? ContextualJudgeBench for Evaluating LLM-based Judges in Contextual Settings](https://mlsanity.com/paper/2503.15620/) (2025)
- [Judging with Many Minds: Do More Perspectives Mean Less Prejudice? On Bias Amplifications and Resistance in Multi-Agent Based LLM-as-Judge](https://mlsanity.com/paper/2505.19477/) (2025)
- [Verify with Caution: The Pitfalls of Relying on Imperfect Factuality Metrics](https://mlsanity.com/paper/2501.14883/) (2025)
- [Quantifying Cognitive Bias Induction in LLM-Generated Content](https://mlsanity.com/paper/2507.03194/) (2025)
- [Hallucinate at the Last in Long Response Generation: A Case Study on Long Document Summarization](https://mlsanity.com/paper/2505.15291/) (2025)
- [QA-prompting: Improving Summarization with Large Language Models using Question-Answering](https://mlsanity.com/paper/2505.14347/) (2025)
- [GenerationPrograms: Fine-grained Attribution with Executable Programs](https://mlsanity.com/paper/2506.14580/) (2025)
- [A Judge Should Know What Changed:Construct Validity for LLM-as-a-Judge Evaluation](https://mlsanity.com/paper/2608.24419/) (2026)
- [StructTest: Benchmarking LLMs' Reasoning through Compositional Structured Outputs](https://mlsanity.com/paper/2412.18011/) (2024)
- [ARC: Argument Representation and Coverage Analysis for Zero-Shot Long Document Summarization with Instruction Following LLMs](https://mlsanity.com/paper/2505.23654/) (2025)
- [Towards a Holistic and Automated Evaluation Framework for Multi-Level Comprehension of LLMs in Book-Length Contexts](https://mlsanity.com/paper/2508.19578/) (2025)
- [Learning from Self Critique and Refinement for Faithful LLM Summarization](https://mlsanity.com/paper/2512.05387/) (2025)
- [Enhancing Long Document Long Form Summarisation with Self-Planning](https://mlsanity.com/paper/2512.17179/) (2025)
- [The Effect of Document Summarization on LLM-Based Relevance Judgments](https://mlsanity.com/paper/2512.05334/) (2025)
- [What Matters to an LLM? Behavioral and Computational Evidences from Summarization](https://mlsanity.com/paper/2602.00459/) (2026)
- [DIAL-SUMMER: A Structured Evaluation Framework of Hierarchical Errors in Dialogue Summaries](https://mlsanity.com/paper/2602.08149/) (2026)
- [A Systematic Evaluation of Positional Bias in Multi-Video Summarization with MLLMs](https://mlsanity.com/paper/2606.04596/) (2026)
- [Summarization is Not Dead Yet](https://mlsanity.com/paper/2606.08000/) (2026)
- [Detect, Remask, Repair: Diffusion Editing for Faithful Summarization of Evolving Contexts](https://mlsanity.com/paper/2606.12807/) (2026)
- [Averaging Bias: Human Faithfulness Annotations are not Locally Faithful](https://mlsanity.com/paper/2608.00205/) (2026)
- [Decoupling Generation and Selection for Budget-Constrained Faithful Summarization](https://mlsanity.com/paper/2608.03655/) (2026)
- [LOOMSUM: Weaving Quantitative and Narrative Evidence for Faithful Long Text-Table Summarization](https://mlsanity.com/paper/2609.00241/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) often exhibit positional bias in long-context settings, under-attending to information in the middle of inputs. We investigate the presence of this bias in long-form summarization, its impact on faithfulness, and various techniques to mitigate this bias. To consistently evaluate faithfulness, we first compile a benchmark of eight human-annotated long-form summarization datasets and perform a meta-evaluation of faithfulness metrics. We show that LLM-based faithfulness metrics, though effective with full-context inputs, remain sensitive to document order, indicating positional bias. Analyzing LLM-generated summaries across six datasets, we find a "U-shaped" trend in faithfulness, where LLMs faithfully summarize the beginning and end of documents but neglect middle content. Perturbing document order similarly reveals models are less faithful when important documents are placed in the middle of the input. We find that this behavior is partly due to shifting focus with context length: as context increases, summaries become less faithful, but beyond a certain length, faithfulness improves as the model focuses on the end. Finally, we experiment with different generation techniques to reduce positional bias and find that prompting techniques effectively direct model attention to specific positions, whereas more sophisticated approaches offer limited improvements. Our data and code are available in https://github.com/meetdavidwan/longformfact.