Future prosaic AIs will likely shape their own development or that of successor AIs. We're trying to make sure they don't go insane.

Latent Diffusion Language Models

Keeping language models honest by directly eliciting knowledge encoded in their activations.

Locating and editing factual associations in GPT (NeurIPS 2022)

petting zoo game theory

Implementation of Perceiver, General Perception with Iterative Attention, in Pytorch