ethav1

3 ms·

1. 1.▲
   by ethav1 3y ago

   It works by self-generating responses to red-team prompts and self-generating safe corrections to those then using RLHF with the corrections. It isn’t a major departure from traditional RLHF so it is interesting that it performs so much bet
2. 2.▲
   by ethav1 4y ago

   Is matching prompts to agents a popular research problem? Does anyone know of some related work?
3. 3.▲
   by ethav1 4y ago

   CUDA applications are usually written in C++. Also, I think HFT normally works with processing tasks that can’t be done in parallel.
4. 4.▲
   by ethav1 4y ago

   Datasets can be manually curated to contain primarily original images if this becomes a real issue. For example, classifiers can predict whether an image is generated or not. You could adapt the process used to create laion-aesthetic[0] to