PrefixNLI: Detecting Factual Inconsistencies as Soon as They Arise

Challenge:	Large Language Models (LLMs) are prone to factually inconsistent statements, known as hallucinations.
Approach:	They propose to train a specialized model that detects inconsistencies over text prefixes to improve generation faithfulness by 5-14 F1 points.
Outcome:	The proposed model outperforms baseline models by 5-14 F1 points in prefix-level entailment.

Similar Papers

Overview

Challenge

Approach

Outcome

Falsesum: Generating Document-level NLI Examples for Recognizing Factual Inconsistency in Summarization (2022.naacl-main)

Copied to clipboard

Prasetya Utama, Joshua Bambrick, Nafise Moosavi, Iryna Gurevych

Information Extraction & Text Mining

Challenge:	Neural abstractive summarization models generate factually inconsistent summaries . previous work has introduced the task of recognizing factual inconsistency as a downstream application of natural language inference (NLI).
Approach:	They propose a data generation pipeline that enables a task-oriented approach to detect factual inconsistencies in abstractive summarization models.
Outcome:	The proposed model improves the state-of-the-art performance across four benchmarks for recognizing factual inconsistency in generated summaries.

Identifying Factual Inconsistencies in Summaries: Grounding LLM Inference via Task Taxonomy (2024.findings-emnlp)

Copied to clipboard

Liyan Xu, Zhenlin Su, Mo Yu, Jin Xu, Jinho Choi, Jie Zhou, Fei Liu

Responsible & Trustworthy NLP

Challenge:	Existing studies have focused on specialized BERT-variants and recent LLMs to reason inconsistencies.
Approach:	They propose to incorporate task-specific taxonomy into inferences to facilitate both zero-shot and supervised paradigms.
Outcome:	The proposed model outperforms specialized non-LLM and recent LLM models in a number of domains.

Detecting Errors through Ensembling Prompts (DEEP): An End-to-End LLM Framework for Detecting Factual Errors (2024.emnlp-main)

Copied to clipboard

Alex Chandler, Devesh Surve, Hui Su

Language Models

Challenge:	Existing methods for detecting factual errors in text summarization are inadequate for the task.
Approach:	They propose an end-to-end large language model framework for detecting factual errors in text summarization.
Outcome:	The proposed framework achieves state-of-the-art (SOTA) balanced accuracy on the AggreFact-XSUM FTSOTA, TofuEval Summary-Level, and HaluEVAL Summarization benchmarks.

SummaC: Re-Visiting NLI-based Models for Inconsistency Detection in Summarization (2022.tacl-1)

Copied to clipboard

Philippe Laban, Tobias Schnabel, Paul N. Bennett, Marti A. Hearst

Summarization

Challenge:	Recent studies have shown that even state-of-the-art pre-trained language models can generate inconsistent summaries in more than 70% of all cases.
Approach:	They propose a method that enables NLI models to be used for inconsistency detection by segmenting documents into sentence units and aggregating scores between pairs of sentences.
Outcome:	The proposed method achieves state-of-the-art accuracy of 74.4% on six large inconsistency detection datasets.

MorphNLI: A Stepwise Approach to Natural Language Inference Using Text Morphing (2025.findings-naacl)

Copied to clipboard

Vlad Andrei Negru, Robert Vacareanu, Camelia Lemnaru, Mihai Surdeanu, Rodica Potolea

Textual Inference

Challenge:	Existing models fail to capture important semantic features of logic such as monotonicity and negation.
Approach:	They propose a modular step-by-step approach to natural language inference . they use a language model to generate edits to incrementally transform the premise into the hypothesis .
Outcome:	The proposed method outperforms baseline models in realistic cross-domain settings with improvements up to 12.6% (relative).

SummEdits: Measuring LLM Ability at Factual Reasoning Through The Lens of Summarization (2023.emnlp-main)

Copied to clipboard

Philippe Laban, Wojciech Kryscinski, Divyansh Agarwal, Alexander Fabbri, Caiming Xiong, Shafiq Joty, Chien-Sheng Wu

Text Generation

Challenge:	Existing factual consistency benchmarks are inadequate to detect factual inconsistencies in LLMs.
Approach:	They propose a protocol for inconsistency detection benchmark creation and implement it in a 10-domain benchmark called SummEdits.
Outcome:	The proposed method is 20 times more cost-effective per sample and highly reproducible, as it estimates inter-annotator agreement at about 0.9.

KNOW How to Make Up Your Mind! Adversarially Detecting and Alleviating Inconsistencies in Natural Language Explanations (2023.acl-short)

Copied to clipboard

Myeongjun Jang, Bodhisattwa Prasad Majumder, Julian McAuley, Thomas Lukasiewicz, Oana-Maria Camburu

Responsible & Trustworthy NLP

Challenge:	eIA is an adversarial attack that generates inconsistent natural language explanations (NLEs) a model that generate In-NLE is undesirable, as it has a faulty decision-making process or is prone to inconsistencies.
Approach:	They propose an off-the-shelf mitigation method to alleviate inconsistencies by grounding the model into external background knowledge.
Outcome:	The proposed method reduces inconsistencies detected by previous models . it is based on external knowledge bases and a novel approach to mitigate inconsistent models based upon the proposed method .

TRUE: Re-evaluating Factual Consistency Evaluation (2022.naacl-main)

Copied to clipboard

Or Honovich, Roee Aharoni, Jonathan Herzig, Hagai Taitelbaum, Doron Kukliansy, Vered Cohen, Thomas Scialom, Idan Szpektor, Avinatan Hassidim, Yossi Matias

Text Generation

Challenge:	Grounded text generation systems often generate factual inconsistencies, hindering their real-world applicability.
Approach:	They propose a method to assess factual consistency metrics on standardized texts . they recommend NLI and question generation-and-answering-based methods as starting points .
Outcome:	The proposed method is more actionable and interpretable than previous methods.

Factually Consistent Summarization via Reinforcement Learning with Textual Entailment Feedback (2023.acl-long)

Copied to clipboard

Paul Roit, Johan Ferret, Lior Shani, Roee Aharoni, Geoffrey Cideron, Robert Dadashi, Matthieu Geist, Sertan Girgin, Leonard Hussenot, Orgad Keller, Nikola Momchev, Sabela Ramos Garea, Piotr Stanczyk, Nino Vieillard, Olivier Bachem, Gal Elidan, Avinatan Hassidim, Olivier Pietquin, Idan Szpektor

Summarization

Challenge:	Recent advances in abstractive summarization systems produce factually inconsistent text . this is emphasized in tasks like summarizing, which often produce inconsistent text with no input article .
Approach:	They use reinforcement learning to optimize for factual consistency and explore trade-offs . they use textual-entailment rewards to optimize the accuracy of the generated summaries .
Outcome:	The proposed method improves faithfulness, salience and conciseness of the generated summaries.

FactKB: Generalizable Factuality Evaluation using Language Models Enhanced with Factual Knowledge (2023.emnlp-main)

Copied to clipboard

Shangbin Feng, Vidhisha Balachandran, Yuyang Bai, Yulia Tsvetkov

Semantic Text Processing

Challenge:	Existing factuality evaluation models are not robust, especially with respect to entity and relation errors in new domains.
Approach:	They propose a new approach to factuality evaluation that is generalizable across domains . they propose entities-specific facts, facts extracted from external knowledge bases and facts constructed compositionally through knowledge base walks.
Outcome:	The proposed model achieves state-of-the-art on two in-domain news summarization benchmarks and on three out-of domain scientific literature datasets.