A survey of learning criteria going beyond the usual risk

MJ Holland, K Tanabe - Journal of Artificial Intelligence Research, 2023 - jair.org
Virtually all machine learning tasks are characterized using some form of loss function, and
“good performance” is typically stated in terms of a sufficiently small average loss, taken over …

Cooperative Inverse Decision Theory for Uncertain Preferences

Z Robertson, H Zhang, S Koyejo - … Conference on Artificial …, 2023 - proceedings.mlr.press
Inverse decision theory (IDT) aims to learn a performance metric for classification by eliciting
expert classifications on examples. However, elicitation in practical settings may require …

Implementability of Information Elicitation Mechanisms with Pre-Trained Language Models

Z Robertson, H Cha, A Sheha, S Koyejo - ICML 2024 Workshop on … - openreview.net
As language models become increasingly sophisticated, ensuring the faithfulness of their
outputs to the input and the consistency of their reasoning across outputs is a critical …