Dataset Groups Activity Stream Self-Supervised Alignment with Mutual Information The dataset is used for training a language model to follow behavioral principles without the use of preference labels, demonstrations, or human oversight. BibTex: @dataset{Jan-Philipp_Fränken_and_Eric_Zelikman_and_Rafael_Rafailov_and_Kanishk_Gandhi_and_Tobias_Gerstenberg_and_Noah_D_Goodman_2024, abstract = {The dataset is used for training a language model to follow behavioral principles without the use of preference labels, demonstrations, or human oversight.}, author = {Jan-Philipp Fränken and Eric Zelikman and Rafael Rafailov and Kanishk Gandhi and Tobias Gerstenberg and Noah D. Goodman}, doi = {10.57702/eafwuidh}, institution = {No Organization}, keyword = {'alignment', 'behavioral principles', 'language model', 'mutual information', 'self-supervised'}, month = {dec}, publisher = {TIB}, title = {Self-Supervised Alignment with Mutual Information}, url = {https://service.tib.eu/ldmservice/dataset/self-supervised-alignment-with-mutual-information}, year = {2024} }