secbench-hf/SecBench

SecBench: A Comprehensive Multi-Dimensional Benchmarking Dataset for LLMs in Cybersecurity 中文README Evaluating Large Language Models (LLMs) is crucial for understanding their capabilities and limitations across various applications, including natural language processing and code generation. Existing benchmarks like MMLU, C-Eval, and HumanEval assess general LLM performance but lack focus on specific expert domains such as cybersecurity. Previous attempts to create cybersecurity… See the full description on the dataset page: https://huggingface.co/datasets/secbench-hf/SecBench.