FABLES: Evaluating faithfulness and content selection in book-length summarization
- URL: http://arxiv.org/abs/2404.01261v1
- Date: Mon, 1 Apr 2024 17:33:38 GMT
- Title: FABLES: Evaluating faithfulness and content selection in book-length summarization
- Authors: Yekyung Kim, Yapei Chang, Marzena Karpinska, Aparna Garimella, Varun Manjunatha, Kyle Lo, Tanya Goyal, Mohit Iyyer,
- Abstract summary: In this paper, we conduct the first large-scale human evaluation of faithfulness and content selection on book-length documents.
We collect FABLES, a dataset of annotations on 3,158 claims made in LLM-generated summaries of 26 books, at a cost of $5.2K USD.
An analysis of the annotations reveals that most unfaithful claims relate to events and character states, and they generally require indirect reasoning over the narrative to invalidate.
- Score: 55.50680057160788
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: While long-context large language models (LLMs) can technically summarize book-length documents (>100K tokens), the length and complexity of the documents have so far prohibited evaluations of input-dependent aspects like faithfulness. In this paper, we conduct the first large-scale human evaluation of faithfulness and content selection on LLM-generated summaries of fictional books. Our study mitigates the issue of data contamination by focusing on summaries of books published in 2023 or 2024, and we hire annotators who have fully read each book prior to the annotation task to minimize cost and cognitive burden. We collect FABLES, a dataset of annotations on 3,158 claims made in LLM-generated summaries of 26 books, at a cost of $5.2K USD, which allows us to rank LLM summarizers based on faithfulness: Claude-3-Opus significantly outperforms all closed-source LLMs, while the open-source Mixtral is on par with GPT-3.5-Turbo. An analysis of the annotations reveals that most unfaithful claims relate to events and character states, and they generally require indirect reasoning over the narrative to invalidate. While LLM-based auto-raters have proven reliable for factuality and coherence in other settings, we implement several LLM raters of faithfulness and find that none correlates strongly with human annotations, especially with regard to detecting unfaithful claims. Our experiments suggest that detecting unfaithful claims is an important future direction not only for summarization evaluation but also as a testbed for long-context understanding. Finally, we move beyond faithfulness by exploring content selection errors in book-length summarization: we develop a typology of omission errors related to crucial narrative elements and also identify a systematic over-emphasis on events occurring towards the end of the book.
Related papers
- Exploring Automatic Cryptographic API Misuse Detection in the Era of LLMs [60.32717556756674]
This paper introduces a systematic evaluation framework to assess Large Language Models in detecting cryptographic misuses.
Our in-depth analysis of 11,940 LLM-generated reports highlights that the inherent instabilities in LLMs can lead to over half of the reports being false positives.
The optimized approach achieves a remarkable detection rate of nearly 90%, surpassing traditional methods and uncovering previously unknown misuses in established benchmarks.
arXiv Detail & Related papers (2024-07-23T15:31:26Z) - STORYSUMM: Evaluating Faithfulness in Story Summarization [31.94902013480574]
We introduce a new dataset, STORYSUMM, comprising short stories with localized faithfulness labels and error explanations.
This benchmark is for evaluation methods, testing whether a given method can detect challenging inconsistencies.
arXiv Detail & Related papers (2024-07-09T02:06:30Z) - WikiContradict: A Benchmark for Evaluating LLMs on Real-World Knowledge Conflicts from Wikipedia [59.96425443250666]
Retrieval-augmented generation (RAG) has emerged as a promising solution to mitigate the limitations of large language models (LLMs)
In this work, we conduct a comprehensive evaluation of LLM-generated answers to questions based on contradictory passages from Wikipedia.
We benchmark a diverse range of both closed and open-source LLMs under different QA scenarios, including RAG with a single passage, and RAG with 2 contradictory passages.
arXiv Detail & Related papers (2024-06-19T20:13:42Z) - TofuEval: Evaluating Hallucinations of LLMs on Topic-Focused Dialogue Summarization [29.49641083851667]
We propose a new evaluation benchmark on topic-focused dialogue summarization, generated by LLMs of varying sizes.
We provide binary sentence-level human annotations of the factual consistency of these summaries along with detailed explanations of factually inconsistent sentences.
arXiv Detail & Related papers (2024-02-20T18:58:49Z) - On Context Utilization in Summarization with Large Language Models [83.84459732796302]
Large language models (LLMs) excel in abstractive summarization tasks, delivering fluent and pertinent summaries.
Recent advancements have extended their capabilities to handle long-input contexts, exceeding 100k tokens.
We conduct the first comprehensive study on context utilization and position bias in summarization.
arXiv Detail & Related papers (2023-10-16T16:45:12Z) - BooookScore: A systematic exploration of book-length summarization in the era of LLMs [53.42917858142565]
We develop an automatic metric, BooookScore, that measures the proportion of sentences in a summary that do not contain any of the identified error types.
We find that closed-source LLMs such as GPT-4 and 2 produce summaries with higher BooookScore than those generated by open-source models.
arXiv Detail & Related papers (2023-10-01T20:46:44Z) - Element-aware Summarization with Large Language Models: Expert-aligned
Evaluation and Chain-of-Thought Method [35.181659789684545]
Automatic summarization generates concise summaries that contain key ideas of source documents.
References from CNN/DailyMail and BBC XSum are noisy, mainly in terms of factual hallucination and information redundancy.
We propose a Summary Chain-of-Thought (SumCoT) technique to elicit LLMs to generate summaries step by step.
Experimental results show our method outperforms state-of-the-art fine-tuned PLMs and zero-shot LLMs by +4.33/+4.77 in ROUGE-L.
arXiv Detail & Related papers (2023-05-22T18:54:35Z) - Evaluating the Factual Consistency of Large Language Models Through News
Summarization [97.04685401448499]
We propose a new benchmark called FIB(Factual Inconsistency Benchmark) that focuses on the task of summarization.
For factually consistent summaries, we use human-written reference summaries that we manually verify as factually consistent.
For factually inconsistent summaries, we generate summaries from a suite of summarization models that we have manually annotated as factually inconsistent.
arXiv Detail & Related papers (2022-11-15T18:50:34Z) - FEQA: A Question Answering Evaluation Framework for Faithfulness
Assessment in Abstractive Summarization [34.2456005415483]
We tackle the problem of evaluating faithfulness of a generated summary given its source document.
We find that current models exhibit a trade-off between abstractiveness and faithfulness.
We propose an automatic question answering (QA) based metric for faithfulness.
arXiv Detail & Related papers (2020-05-07T21:00:08Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.