pplx-pii-masking

Discussed in 1 episode

Episodes

#5481: Small Models, Big Guardrails: PII Detection in ChatGPT

A tiny 600M parameter model is quietly scanning your ChatGPT tool calls for PII. Here's how that class of guardian model actually works.

Main topic Sep 23