Chinese SafetyQA: A Safety Short-form Factuality Benchmark for Large Language Models

TL;DR

This paper introduces Chinese SafetyQA, a benchmark for evaluating the safety-related factuality of large language models in Chinese, focusing on accuracy, comprehensiveness, and safety in sensitive domains.

Contribution

It presents a new Chinese safety factuality benchmark and evaluates existing LLMs' capabilities, analyzing their safety knowledge accuracy and robustness.

Findings

Existing LLMs show varying safety factuality performance.

Chinese SafetyQA is effective for evaluating safety-related factuality.

Analysis reveals gaps in LLM safety knowledge understanding.

Abstract

With the rapid advancement of Large Language Models (LLMs), significant safety concerns have emerged. Fundamentally, the safety of large language models is closely linked to the accuracy, comprehensiveness, and clarity of their understanding of safety knowledge, particularly in domains such as law, policy and ethics. This factuality ability is crucial in determining whether these models can be deployed and applied safely and compliantly within specific regions. To address these challenges and better evaluate the factuality ability of LLMs to answer short questions, we introduce the Chinese SafetyQA benchmark. Chinese SafetyQA has several properties (i.e., Chinese, Diverse, High-quality, Static, Easy-to-evaluate, Safety-related, Harmless). Based on Chinese SafetyQA, we perform a comprehensive evaluation on the factuality abilities of existing LLMs and analyze how these capabilities relate to LLM abilities, e.g., RAG ability and robustness against attacks.