Abstract
In spoken languages, utterances are often shaped to be incomplete or vague for efficiency. This can lead to varying interpretations of the same input, based on different assumptions about the context. To ensure reliable user-model interactions in such scenarios, it is crucial for models to adeptly handle the inherent ambiguity in user queries. However, conversational agents built upon even the most recent large language models (LLMs) face challenges in processing ambiguous inputs, primarily due to the following two hurdles: (1) LLMs are not directly trained to handle inputs that are too ambiguous to be properly managed; (2) the degree of ambiguity in an input can vary according to the intrinsic knowledge of the LLMs, which is difficult to investigate. To address these issues, this paper proposes a method to align LLMs to explicitly handle ambiguous inputs. Specifically, we introduce a proxy task that guides LLMs to utilize their intrinsic knowledge to self-disambiguate a given input. We quantify the information gain from the disambiguation procedure as a measure of the extent to which the models perceive their inputs as ambiguous. This measure serves as a cue for selecting samples deemed ambiguous from the models' perspectives, which are then utilized for alignment. Experimental results from several question-answering datasets demonstrate that the LLMs fine-tuned with our approach are capable of handling ambiguous inputs while still performing competitively on clear questions within the task.
Abstract (translated)
在口语中，常常需要让陈述不完整或含糊不清，以提高效率。这可能导致对相同输入的不同假设导致对相同输入的不同解释。为了在类似情况下确保可靠的用户-模型交互，模型需要巧妙处理用户查询固有的歧义性。然而，基于最新的大型语言模型（LLM）构建的会话机器人面临处理模糊输入的挑战，主要原因有以下两个障碍：（1）LLM 没有直接训练来处理过于复杂或无法正确管理的输入；（2）LLM 固有的知识可能随其内在知识而变化，这很难进行调查。为了应对这些问题，本文提出了一种将 LLM 对齐为明确处理模糊输入的方法。具体来说，我们引入了一个引导任务，使 LLM 使用其固有知识来自我消除给定的输入。我们衡量消除过程的信息增益作为衡量模型认为其输入存在歧义的程度。这个度量作为从模型观点选择认为存在歧义的样本的提示。在多个问题回答数据集的实验结果中，采用我们方法训练的 LLM 能够处理模糊输入，同时在任务内保持竞争力。
URL
https://arxiv.org/abs/2404.11972
