Confiance AI ; Tweet AND Tweet
Common descendants
1 Document
Anthropic sur Twitter : "We examine which safety techniques for LMs are more robust to human-written, adversarial inputs ..."
Tags:
2022-08-25 About