Gradient Defense - Can you simply brainwash an LLM?
Raises concerns about traceability and potential manipulations in the supply chain of large language models. We explore a proposed solution to address these issues.
Source: gradientdefense.com