We present an exploratory framework to test whether noise-like input can induce structured responses in language models.
We present an exploratory framework to test whether noise-like input can induce structured responses in language models.