Survey of Hallucination in Natural Language Generation

Abstract

Natural Language Generation (NLG) has improved exponentially in recent years thanks to the development of sequence-to-sequence deep learning technologies such as Transformer-based language models. This advancement has led to more fluent and coherent NLG, leading to improved development in downstream tasks such as abstractive summarization, dialogue generation and data-to-text generation. However, it is also apparent that deep learning based generation is prone to hallucinate unintended text, which degrades the system performance and fails to meet user expectations in many real-world scenarios. To address this issue, many studies have been presented in measuring and mitigating hallucinated texts, but these have never been reviewed in a comprehensive manner before. In this survey, we thus provide a broad overview of the research progress and challenges in the hallucination problem in NLG. The survey is organized into two parts: (1) a general overview of metrics, mitigation methods, and future directions; and (2) an overview of task-specific research progress on hallucinations in the following downstream tasks, namely abstractive summarization, dialogue generation, generative question answering, data-to-text generation, and machine translation. This survey serves to facilitate collaborative efforts among researchers in tackling the challenge of hallucinated texts in NLG.

Human-level play in the game of Diplomacy by combining language models with strategic reasoning

“…Prior work has shown that neural language models suffer from contradictions and inconsistency as well as a tendency to “hallucinate,” or generate factually incorrect information ( 29 ). In the complex domain of Diplomacy , dialogue models exhibit both these problems and other more subtle mistakes, such as deviations from the intents used to control the message or blunders in the strategic content of the message.…”

Gpt as Knowledge Worker: A Zero-Shot Evaluation of (AI)CPA Capabilities

“…Out of all explanations, including incorrect ones, explanations included at least one hallucinated reference or authority in approximately 37% of the time. Research is ongoing on the optimal degree of hallucination and techniques for mitigating unwanted hallucination [38], and we will continue to explore these questions and applications in future work. For text-davinci-003, the average is reported across all runs; for other models, a subset of representative prompts and parameters were included.…”

Navigating Complexity in Software Engineering: A Prototype for Comparing GPT-n Solutions

“…Although these natural language processing techniques have the potential to greatly improve the efficiency and effectiveness of software development, there are concerns regarding their use. One problem with large language models is that they can sometimes "hallucinate" [24]. A recent study found that a state-of-the-art model was more likely to generate code containing a vulnerability if the query asked for code without that vulnerability [25].…”