leondz/lm_risk_cards
Risks and targets for assessing LLMs & LLM vulnerabilities
These Language Model Risk Cards help you systematically identify potential problems and vulnerabilities in how you plan to use a large language model. You'll choose a specific use case, model, and interface, then select relevant cards to guide your testing. The outcome is a detailed assessment report based on your efforts to provoke specific risky behaviors from the LLM. This is for anyone responsible for safely and effectively deploying large language models, such as product managers, AI ethics officers, or compliance specialists.
No commits in the last 6 months.
Use this if you need a structured way to uncover potential failures or risks before deploying a large language model into a real-world application.
Not ideal if you are looking for an automated tool to run tests or a technical library for developers to integrate into their code.
Stars
34
Forks
10
Language
Python
License
—
Category
Last pushed
May 27, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/leondz/lm_risk_cards"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related models
zealscott/AutoProfiler
Source code for Automated Profile Inference with Language Model Agents
shreyansh26/Red-Teaming-Language-Models-with-Language-Models
A re-implementation of the "Red Teaming Language Models with Language Models" paper by Perez et al., 2022
RedTeamingforLLMs/RedTeamingforLLMs
A framework designed for executing positive red-teaming experiments on large language models.
dan0nchik/llm-attack-kit
A collection of LLM attacks