| """ |
| Text content for the GuardBench Leaderboard. |
| """ |
|
|
| TITLE = """ |
| <div style="text-align: center; margin-bottom: 1rem"> |
| <h1>GuardBench Leaderboard</h1> |
| </div> |
| """ |
|
|
| INTRODUCTION_TEXT = """ |
| ## Introduction |
| |
| GuardBench is a comprehensive benchmark for evaluating the safety guardrails of large language models (LLMs). |
| This leaderboard tracks model performance across various safety categories, including harmful content detection, |
| jailbreak resistance, and more. |
| |
| Models are evaluated on their ability to properly refuse harmful requests and detect problematic content |
| across multiple categories and test scenarios. |
| """ |
|
|
| LLM_BENCHMARKS_TEXT = """ |
| GuardBench checks how well models handle safety challenges — from misinformation and self-harm to sexual content and corruption. |
| |
| Models are tested with regular and adversarial prompts to see if they can avoid saying harmful things. |
| |
| We track how accurate they are, how often they make mistakes, and how fast they respond. |
| """ |
|
|
| EVALUATION_QUEUE_TEXT = """ |
| ## Submit Your Model |
| |
| To add your model to the GuardBench leaderboard: |
| |
| Run your evaluation using the GuardBench framework at https://github.com/whitecircle-ai/guard-bench |
| Upload your run results in .jsonl format using this form |
| Once validated, your model will appear on the leaderboard. |
| |
| ✉️✨ Ready? Upload your results below! |
| """ |
|
|
| CITATION_BUTTON_LABEL = "Cite GuardBench" |
|
|
| CITATION_BUTTON_TEXT = """ |
| @misc{guardbench2023, |
| author = {GuardBench Team}, |
| title = {GuardBench: Comprehensive Benchmark for LLM Safety Guardrails}, |
| year = {2023}, |
| publisher = {GitHub}, |
| journal = {GitHub repository}, |
| howpublished = {\\url{https://github.com/huggingface/guard-bench}} |
| } |
| """ |
|
|
