In this paper, we proposed an efficient weak-supervised adversarial data augmentation approach to form the factual consistency dataset.

Nov 11, 2021 · In this paper, we propose a robust weak- supervised factual consistency evaluation model and gradient-based factual errors tracing strategy.

In this paper, we propose a robust weak- supervised factual consistency evaluation model and gradient-based factual errors tracing strategy. Specifically, we ...

Gradient-Based Adversarial Factual Consistency Evaluation for Abstractive Summarization ... The content you want is available to Zendy users.Already have an ...

Connected Papers is a visual tool to help researchers and applied scientists find academic papers relevant to their field of work.

Gradient-Based Adversarial Factual Consistency Evaluation for Abstractive Summarization ... Neural abstractive summarization systems have gained significant ...

Based on the artificial dataset, we train an evaluation model that can not only make accurate and robust factual consistency discrimination but is also capable ...

May 26, 2023 · The propensity of abstractive summarization models to make factual errors has been studied extensively, including design of metrics to de-.

JiazeChen. Latest. Gradient-Based Adversarial Factual Consistency Evaluation for Abstractive Summarization. © PRIS-NLP 2024. Published with Wowchemy Website ...

In this paper, we propose a simple but strong energy-based score function to detect OOD where the energy scores of OOD samples are higher than IND samples.