Models exhibit covert value leakage: the information they provide is influenced by their own values, without this influence being disclosed to the user. We introduce a suite of evaluations to quantify value leakage and find large differences among frontier models.

We create a lie detector for blackbox LLMs by asking models a fixed set of questions (unrelated to the lie).

We show that a GPT-3 model can learn to express uncertainty about its own answers in natural language -- without use of model logits.

We propose a benchmark to measure whether a language model is truthful in generating answers to questions.