Lately, there seems to be a trend in natural language models to have some sort of knowledge base lookup or memory. I'm guessing it's only a matter of time unti

Lately, there seems to be a trend in natural language models to have some sort of knowledge base lookup or memory. I'm guessing it's only a matter of time until this comes to toxicity models, so they can look up who said it, and to who it was said.