https://github.com/LLM-Tuning-Safety/LLMs-Finetuning-Safety

We jailbreak GPT-3.5 Turbo’s safety guardrails by fine-tuning it on only 10 adversarially designed examples, at a cost of less than $0.20 via OpenAI’s APIs.
https://github.com/LLM-Tuning-Safety/LLMs-Finetuning-Safety

alignment llm llm-finetuning

Last synced: 10 days ago
JSON representation

We jailbreak GPT-3.5 Turbo’s safety guardrails by fine-tuning it on only 10 adversarially designed examples, at a cost of less than $0.20 via OpenAI’s APIs.

• Host: GitHub
• URL: https://github.com/LLM-Tuning-Safety/LLMs-Finetuning-Safety
• Owner: LLM-Tuning-Safety
• License: mit
• Created: 2023-10-06T16:02:27.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-02-23T21:19:44.000Z (10 months ago)
• Last Synced: 2024-08-12T08:09:31.085Z (4 months ago)
• Topics: alignment, llm, llm-finetuning
• Language: Python
• Homepage: https://llm-tuning-safety.github.io/
• Size: 23.2 MB
• Stars: 211
• Watchers: 4
• Forks: 21
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• Awesome-MLLM-Safety - Github - Tuning-Safety/LLMs-Finetuning-Safety.svg?style=social&label=Star) (Evaluation)
• awesome-MLSecOps - LLMs-Finetuning-Safety
• Awesome-LLMSecOps - LLMs Finetuning Safety - tuning large language models |  | (PoC)