More Like this

1. Aligning to Adults Is Easy, Aligning to Children Is Hard: A Study of Linguistic Alignment in Dialogue Systems

   https://doi.org/10.18653/v1/2024.hucllm-1.7  

   French, Dorothea; D’Mello, Sidney; von_der_Wense, Katharina (January 2024, ACL)
2. Aligning Patterns to the Wikibase Model

   Eels, Andrew; Shimizu, Cogan; Zhou, Lu; Hitzler, Pascal; Gonzalez, Seila; Rehberger, Dean (January 2021, WOP 2021 Workshop on Ontology Design and Patterns 2021 Proceedings of the 12th Workshop on Ontology Design and Patterns (WOP 2021) co-located with the 20th International Semantic Web Conference (ISWC 2021) Online, October 24, 2021.)
3. Aligning data with decisions to address the biodiversity crisis

   https://doi.org/10.1371/journal.pbio.3002683  

   Gerber, Leah R; Iacona, Gwenllian D (June 2024, PLOS Biology)
4. Aligning an optical cavity: with reference to cavity ring-down spectroscopy

   https://doi.org/10.1364/AO.405189  

   Telfah, Hamzeh; Paul, Anam_C; Liu, Jinjun (October 2020, Applied Optics)

   A procedure for timely, accurate, and reproducible alignment of an optical cavity is described. 

   more » « less
5. A Dense Reward View on Aligning Text-to-Image Diffusion with Preference

   Yang, Shentao; Chen, Tianqi; Zhou, Mingyuan (July 2024, International Conference on Machine Learning)

   Aligning text-to-image diffusion model (T2I) with preference has been gaining increasing research attention. While prior works exist on directly optimizing T2I by preference data, these methods are developed under the bandit assumption of a latent reward on the entire diffusion reverse chain, while ignoring the sequential nature of the generation process. From literature, this may harm the efficacy and efficiency of alignment. In this paper, we take on a finer dense reward perspective and derive a tractable alignment objective that emphasizes the initial steps of the T2I reverse chain. In particular, we introduce temporal discounting into the DPO-style explicit-reward-free loss, to break the temporal symmetry therein and suit the T2I generation hierarchy. In experiments on single and multiple prompt generation, our method is competitive with strong relevant baselines, both quantitatively and qualitatively. Further studies are conducted to illustrate the insight of our approach. 

   more » « less