Goto

Collaborating Authors

 Madison County


Dementia risk could increase with low levels of essential vitamin

FOX News

Fox News contributor Dr. Marc Siegel joins'Fox News Live' to discuss the FDA approving a new Alzheimer treatment drug and the FDA banning bromide vegetable oils. "Normal" levels of vitamin B12 may not be enough to ward off dementia, new research finds. Researchers at University of California San Francisco studied 231 healthy older adults (averaging 71 years of age) who did not have dementia or mild cognitive impairment. Blood tests showed that their B12 levels averaged 414.8 pmol/L, while the recommended minimum level in the U.S. is just 148 pmol/L. Participants who had lower B12 levels were found to have "slower cognitive and visual processing speeds" when taking tests, which is linked to "subtle cognitive decline," according to a UCSF press release.


Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View

Liu, Tianyu, Zheng, Xin, Chang, Baobao, Sui, Zhifang

arXiv.org Artificial Intelligence

In open domain table-to-text generation, we notice that the unfaithful generation usually contains hallucinated content which can not be aligned to any input table record. We thus try to evaluate the generation faithfulness with two entity-centric metrics: table record coverage and the ratio of hallucinated entities in text, both of which are shown to have strong agreement with human judgements. Then based on these metrics, we quantitatively analyze the correlation between training data quality and generation fidelity which indicates the potential usage of entity information in faithful generation. Motivated by these findings, we propose two methods for faithful generation: 1) augmented training by incorporating the auxiliary entity information, including both an augmented plan-based model and an unsupervised model and 2) training instance selection based on faithfulness ranking. We show these approaches improve generation fidelity in both full dataset setting and few shot learning settings by both automatic and human evaluations.