中文
相关论文

相关论文: The Curious Case of Neural Text Degeneration

200 篇论文

Neural text generation is a key tool in natural language applications, but it is well known there are major problems at its core. In particular, standard likelihood training and decoding leads to dull and repetitive outputs. While some…

机器学习 · 计算机科学 2019-09-30 Sean Welleck , Ilia Kulikov , Stephen Roller , Emily Dinan , Kyunghyun Cho , Jason Weston

Despite their growing capabilities, language models still frequently reproduce content from their training data, generate repetitive text, and favor common grammatical patterns and vocabulary. A possible cause is the decoding strategy: the…

计算与语言 · 计算机科学 2026-01-15 Giorgio Franceschelli , Mirco Musolesi

Recent advancements in neural language modelling make it possible to rapidly generate vast amounts of human-sounding text. The capabilities of humans and automatic discriminators to detect machine-generated text have been a large source of…

计算与语言 · 计算机科学 2020-05-11 Daphne Ippolito , Daniel Duckworth , Chris Callison-Burch , Douglas Eck

When generating text from probabilistic models, the chosen decoding strategy has a profound effect on the resulting text. Yet the properties elicited by various decoding strategies do not always transfer across natural language generation…

计算与语言 · 计算机科学 2022-03-30 Gian Wiher , Clara Meister , Ryan Cotterell

This work analyses the text memorization behavior of large language models (LLMs) when subjected to nucleus sampling. Stochastic decoding methods like nucleus sampling are typically applied to overcome issues such as monotonous and…

计算与语言 · 计算机科学 2024-08-30 Luka Borec , Philipp Sadler , David Schlangen

There are a number of diverging hypotheses about the neural text degeneration problem, i.e., generating repetitive and dull loops, which makes this problem both interesting and confusing. In this work, we aim to advance our understanding by…

计算与语言 · 计算机科学 2023-10-17 Huayang Li , Tian Lan , Zihao Fu , Deng Cai , Lemao Liu , Nigel Collier , Taro Watanabe , Yixuan Su

Neural language models often fail to generate diverse and informative texts, limiting their applicability in real-world problems. While previous approaches have proposed to address these issues by identifying and penalizing undesirable…

计算与语言 · 计算机科学 2023-09-25 Jimin Hong , ChaeHun Park , Jaegul Choo

Deep learning methods have recently achieved great empirical success on machine translation, dialogue response generation, summarization, and other text generation tasks. At a high level, the technique has been to train end-to-end neural…

计算与语言 · 计算机科学 2017-11-28 Ziang Xie

Likelihood training and maximization-based decoding result in dull and repetitive generated texts even when using powerful language models (Holtzman et al., 2019). Adding a loss function for regularization was shown to improve text…

计算与语言 · 计算机科学 2021-01-13 Evgeny Lagutin , Daniil Gavrilov , Pavel Kalaidin

As Large Language Models (LLMs) become increasingly prevalent, their generated outputs are proliferating across the web, risking a future where machine-generated content dilutes human-authored text. Since online data is the primary resource…

计算与语言 · 计算机科学 2025-09-23 George Drayson , Emine Yilmaz , Vasileios Lampos

The neural text generation suffers from the text degeneration issue such as repetition. Traditional stochastic sampling methods only focus on truncating the unreliable "tail" of the distribution, and do not address the "head" part, which we…

计算与语言 · 计算机科学 2021-08-27 Xinran Zhang , Maosong Sun , Jiafeng Liu , Xiaobing Li

Despite their ubiquity in language generation, it remains unknown why truncation sampling heuristics like nucleus sampling are so effective. We provide a theoretical explanation for the effectiveness of the truncation sampling by proving…

计算与语言 · 计算机科学 2023-10-04 Matthew Finlayson , John Hewitt , Alexander Koller , Swabha Swayamdipta , Ashish Sabharwal

Standard decoding strategies for text generation, including top-k, nucleus sampling, and contrastive search, select tokens based on likelihood, restricting selection to high-probability regions. Human language production operates…

计算与语言 · 计算机科学 2026-03-20 Esteban Garces Arias , Nurzhan Sapargali , Christian Heumann , Matthias Aßenmacher

Text generation is of great importance to many natural language processing applications. However, maximization-based decoding methods (e.g. beam search) of neural language models often lead to degenerate solutions -- the generated text is…

计算与语言 · 计算机科学 2022-09-27 Yixuan Su , Tian Lan , Yan Wang , Dani Yogatama , Lingpeng Kong , Nigel Collier

For open-ended language generation tasks such as storytelling and dialogue, choosing the right decoding algorithm is critical to controlling the tradeoff between generation quality and diversity. However, there presently exists no consensus…

计算与语言 · 计算机科学 2020-04-23 Hugh Zhang , Daniel Duckworth , Daphne Ippolito , Arvind Neelakantan

Language models suffer from various degenerate behaviors. These differ between tasks: machine translation (MT) exhibits length bias, while tasks like story generation exhibit excessive repetition. Recent work has attributed the difference…

计算与语言 · 计算机科学 2022-10-21 Darcey Riley , David Chiang

Today's probabilistic language generators fall short when it comes to producing coherent and fluent text despite the fact that the underlying models perform well under standard metrics, e.g., perplexity. This discrepancy has puzzled the…

计算与语言 · 计算机科学 2025-06-06 Clara Meister , Tiago Pimentel , Gian Wiher , Ryan Cotterell

Despite the remarkable advances in language modeling, current mainstream decoding methods still struggle to generate texts that align with human texts across different aspects. In particular, sampling-based methods produce less-repetitive…

计算与语言 · 计算机科学 2024-06-06 Haozhe Ji , Pei Ke , Hongning Wang , Minlie Huang

This work focuses on relating two mysteries in neural-based text generation: exposure bias, and text degeneration. Despite the long time since exposure bias was mentioned and the numerous studies for its remedy, to our knowledge, its impact…

计算与语言 · 计算机科学 2021-09-21 Ting-Rui Chiang , Yun-Nung Chen

In Natural Language Generation (NLG) tasks, for any input, multiple communicative goals are plausible, and any goal can be put into words, or produced, in multiple ways. We characterise the extent to which human production varies lexically,…

计算与语言 · 计算机科学 2023-10-23 Mario Giulianelli , Joris Baan , Wilker Aziz , Raquel Fernández , Barbara Plank
‹ 上一页 1 2 3 10 下一页 ›