Linguistic calibration through metacognition: aligning dialogue agent responses with expected correctness

12/30/2020 ∙ by Sabrina J. Mielke, et al. ∙ 0

Open-domain dialogue agents have vastly improved, but still confidently hallucinate knowledge or express doubt when asked straightforward questions. In this work, we analyze whether state-of-the-art chit-chat models can express metacognition capabilities through their responses: does a verbalized expression of doubt (or confidence) match the likelihood that the model's answer is incorrect (or correct)? We find that these models are poorly calibrated in this sense, yet we show that the representations within the models can be used to accurately predict likelihood of correctness. By incorporating these correctness predictions into the training of a controllable generation model, we obtain a dialogue agent with greatly improved linguistic calibration.



There are no comments yet.


page 13

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.