Safety-Prompts
Safety data set
Provides a dataset of safety prompts to evaluate and improve the safety of large language models.
Chinese safety prompts for evaluating and improving the safety of LLMs. 中文安全prompts，用于评估和提升大模型的安全性。
880 stars
7 watching
81 forks
last commit: 12 months ago
Linked from 1 awesome list
attack-defensechatgptchinese-languageinstructionllmpromptprompt-engineeringsafety
yaodongc/awesome-instruction-datasetRelated projects:
| Repository | Description | Stars |
|---|---|---|
| thu-coai/cdial-gpt | A large-scale Chinese conversation dataset and pre-trained dialog models for text generation | 1,799 |
| thu-coai/opd | A large-scale pre-trained dialogue model for Chinese language | 74 |
| thu-coai/eva | Pre-trained chatbot models for Chinese open-domain dialogue systems | 306 |
 travistangvh/chatgpt-data-science-prompts | A collection of data science prompts for ChatGPT, providing a set of curated input text to generate human-like responses on various machine learning and data-related topics. | 1,470 |
 x-plug/cvalues | Evaluates and aligns the values of Chinese large language models with safety and responsibility standards | 481 |
 ys-zong/vlguard | Improves safety and helpfulness of large language models by fine-tuning them using safety-critical tasks | 47 |
 logpai/loghub | Provides a collection of system log datasets for AI-driven analytics research. | 1,883 |
 chatopera/insuranceqa-corpus-zh | An insurance industry conversation corpus with pre-processed data for natural language processing and question answering tasks. | 1,019 |
 protectai/rebuff | Protects AI applications from prompt injection attacks through multiple layers of defense | 1,144 |
 deadbits/vigil-llm | A security scanner for Large Language Model prompts to detect potential threats and vulnerabilities | 326 |
 udacity/carnd-functional-safety-project | Developing a comprehensive documentation set for a lane assistance system to ensure its functional safety | 84 |
 michael-wzhu/promptcblue | A large-scale instruction-tuning dataset for multi-task and few-shot learning in the medical domain | 328 |
 xuefuzhao/instructionwild | Creating a large-scale user-based instruction dataset for natural language processing research and development | 455 |
 brightmart/xlnet_zh | Trains a large Chinese language model on massive data and provides a pre-trained model for downstream tasks | 230 |
 commaai/panda | A software system providing a safety-critical interface to autonomous vehicles' CAN bus networks | 1,544 |