# CHiSafetyBench: A Chinese Hierarchical Safety Benchmark for Large Language Models

arXiv: https://arxiv.org/abs/2406.10311 · HTML: https://mlsanity.com/paper/2406.10311/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 27 references that are in the map.
- First posted on arXiv 14 Jun 2024.

## Builds on (most cited first, 24 of 24)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection](https://mlsanity.com/paper/2203.09509/) (2022)
- [BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation](https://mlsanity.com/paper/2101.11718/) (2021)
- [Multilingual Jailbreak Challenges in Large Language Models](https://mlsanity.com/paper/2310.06474/) (2023)
- [ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation](https://mlsanity.com/paper/2310.17389/) (2023)
- [SALAD-Bench: A Hierarchical and Comprehensive Safety Benchmark for Large Language Models](https://mlsanity.com/paper/2402.05044/) (2024)
- [Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignment](https://mlsanity.com/paper/2308.09662/) (2023)
- [SafetyBench: Evaluating the Safety of Large Language Models](https://mlsanity.com/paper/2309.07045/) (2023)
- [Large Language Models for Generative Information Extraction: A Survey](https://mlsanity.com/paper/2312.17617/) (2023)
- [Attacks, Defenses and Evaluations for LLM Conversation Safety: A Survey](https://mlsanity.com/paper/2402.09283/) (2024)
- [A Survey of Safety and Trustworthiness of Large Language Models through the Lens of Verification and Validation](https://mlsanity.com/paper/2305.11391/) (2023)
- [Safety Assessment of Chinese Large Language Models](https://mlsanity.com/paper/2304.10436/) (2023)
- [CValues: Measuring the Values of Chinese Large Language Models from Safety to Responsibility](https://mlsanity.com/paper/2307.09705/) (2023)
- [All Languages Matter: On the Multilingual Safety of Large Language Models](https://mlsanity.com/paper/2310.00905/) (2023)
- [SC-Safety: A Multi-round Open-ended Question Adversarial Safety Benchmark for Large Language Models in Chinese](https://mlsanity.com/paper/2310.05818/) (2023)
- [Fair Enough: Standardizing Evaluation and Model Selection for Fairness Research in NLP](https://mlsanity.com/paper/2302.05711/) (2023)
- [Systematic Evaluation of Predictive Fairness](https://mlsanity.com/paper/2210.08758/) (2022)
- [Large Language Models in Targeted Sentiment Analysis](https://mlsanity.com/paper/2404.12342/) (2024)

## Built on by (most cited first, 15 of 15)

- [PolyGuard: A Multilingual Safety Moderation Tool for 17 Languages](https://mlsanity.com/paper/2504.04377/) (2025)
- [Toward Generalizable Evaluation in the LLM Era: A Survey Beyond Benchmarks](https://mlsanity.com/paper/2504.18838/) (2025)
- [The Scales of Justitia: A Comprehensive Survey on Safety Evaluation of LLMs](https://mlsanity.com/paper/2506.11094/) (2025)
- [A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Models](https://mlsanity.com/paper/2509.03871/) (2025)
- [o3-mini vs DeepSeek-R1: Which One is Safer?](https://mlsanity.com/paper/2501.18438/) (2025)
- [Early External Safety Testing of OpenAI's o3-mini: Insights from the Pre-Deployment Evaluation](https://mlsanity.com/paper/2501.17749/) (2025)
- [Safety Evaluation of DeepSeek Models in Chinese Contexts](https://mlsanity.com/paper/2502.11137/) (2025)
- [ChineseSafe: A Chinese Benchmark for Evaluating Safety in Large Language Models](https://mlsanity.com/paper/2410.18491/) (2024)
- [ASTRAL: Automated Safety Testing of Large Language Models](https://mlsanity.com/paper/2501.17132/) (2025)
- [Methodology of Adapting Large English Language Models for Specific Cultural Contexts](https://mlsanity.com/paper/2406.18192/) (2024)
- [Safety Evaluation and Enhancement of DeepSeek Models in Chinese Contexts](https://mlsanity.com/paper/2503.16529/) (2025)
- [Ethical Risks in Deploying Large Language Models: An Evaluation of Medical Ethics Jailbreaking](https://mlsanity.com/paper/2601.12652/) (2026)
- [Yesterday's Shield, Today's Spear: A Self-Evolving Safety Guardrail in Production](https://mlsanity.com/paper/2608.08471/) (2026)
- [Who Judges the Judges? A Chinese Safety QA Benchmark for Evaluating LLM Responses and Safety Judges](https://mlsanity.com/paper/2609.01210/) (2026)
- [PUBG Ally: A Conversational Embodied Agent as an AI Teammate](https://mlsanity.com/paper/2609.29837/) (2026)

## Abstract (arXiv)

With the profound development of large language models(LLMs), their safety concerns have garnered increasing attention. However, there is a scarcity of Chinese safety benchmarks for LLMs, and the existing safety taxonomies are inadequate, lacking comprehensive safety detection capabilities in authentic Chinese scenarios. In this work, we introduce CHiSafetyBench, a dedicated safety benchmark for evaluating LLMs' capabilities in identifying risky content and refusing answering risky questions in Chinese contexts. CHiSafetyBench incorporates a dataset that covers a hierarchical Chinese safety taxonomy consisting of 5 risk areas and 31 categories. This dataset comprises two types of tasks: multiple-choice questions and question-answering, evaluating LLMs from the perspectives of risk content identification and the ability to refuse answering risky questions respectively. Utilizing this benchmark, we validate the feasibility of automatic evaluation as a substitute for human evaluation and conduct comprehensive automatic safety assessments on mainstream Chinese LLMs. Our experiments reveal the varying performance of different models across various safety domains, indicating that all models possess considerable potential for improvement in Chinese safety capabilities. Our dataset is publicly available at https://github.com/UnicomAI/UnicomBenchmark/tree/main/CHiSafetyBench.