提出ObfusQAte框架，测试大模型在伪装问题下的问答鲁棒性。
ObfusQAte: A Proposed Framework to Evaluate LLM Robustness on Obfuscated Factual Question Answering
- 设计多层级伪装策略，模拟实体误导、干扰项和上下文过载。
- 发现大模型在复杂伪装问题下错误率上升，常出现幻觉回答。
- 适合研究模型鲁棒性与对抗攻击的学者使用。
大型语言模型（LLMs）在事实型问答任务中表现优异，但现有研究未系统评估其在问题被伪装时的鲁棒性。为此，我们提出ObfusQAte方法，并构建首个综合性框架ObfusQA，包含三层伪装机制：(i) 实体间接指代，(ii) 干扰项间接指代，(iii) 上下文过载。该框架从三个维度全面评估模型适应能力。实验表明，当问题经过上述伪装后，大模型倾向于失败或生成幻觉答案。为推动该方向研究，我们公开发布ObfusQAte工具包。
原文摘要 · Abstract (English)
The rapid proliferation of Large Language Models (LLMs) has significantly contributed to the development of equitable AI systems capable of factual question-answering (QA). However, no known study tests the LLMs' robustness when presented with obfuscated versions of questions. To systematically evaluate these limitations, we propose a novel technique, ObfusQAte, and leveraging the same, introduce ObfusQA, a comprehensive, first-of-its-kind framework with multi-tiered obfuscation levels designed to examine LLM capabilities across three distinct dimensions: (i) Named-Entity Indirection, (ii) Distractor Indirection, and (iii) Contextual Overload. By capturing these fine-grained distinctions in language, ObfusQA provides a comprehensive benchmark for evaluating LLM robustness and adaptability. Our study observes that LLMs exhibit a tendency to fail or generate hallucinated responses when confronted with these increasingly nuanced variations. To foster research in this direction, we make ObfusQAte publicly available.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。