Search

Advancements in deep neural networks have led to significant progress in computer vision and natural language processing. These networks, trained on real-world stimuli, develop high-level feature representations of stimuli. It is hypothesized that these representations, stemming from different inputs, should converge into similar conceptual systems, as they reflect various perspectives of the same underlying reality. This paper examines the degree to which different conceptual systems can be aligned in an unsupervised manner, using feature-based representations from deep neural networks. Our investigation centers on the alignment between the image and word representations produced by diverse neural networks, emphasizing those trained via self-supervised learning methods. Subsequently, to probe comparable alignment patterns in human learning, we extend this examination to models trained on developmental headcam data from children. Our findings reveal a more pronounced alignment in models trained through self-supervised learning compared to supervised learning, effectively uncovering higher-level structural connections among categories. However, this alignment was notably absent in models trained with limited developmental headcam data, suggesting more data, more inductive biases, or more supervision are needed to establish alignment from realistic input.

The development of edible vegetable oil is related to food security and people’s quality of life. Literature research method was adopted in this research to analyze the policy and economic environments of edible vegetable oil industry, and on this basis, production scale, sales quantity and brand competition of edible vegetable oil were sorted out. According to the production data of edible vegetable oil from 2013 to 2022, the GM (1,1) model was used to forecast the yields of edible vegetable oil in the next five years. The results revealed that the output of edible vegetable oil will decline in the future. There was a super-large market advantage in the domestic market, and the overall development of edible vegetable oil industry will be stable in the future. It is concluded that the edible vegetable oil industry may face market challenges such as consumer demand reduction, demand structure diversification and complex competition of market players in the future. The study suggested that, first, enterprises need to pay attention to the nutrition and health of edible oil products, which could ensure food quality and safety, and meet the needs of consumers for a healthy life. Second, enterprises should carry out product innovation to develop new oil products with high quality as well as lightweight, and create a new consumption scene. Third, by targeting the niche consumer market, actively expanding the market space of edible oil and stabilizing the market scale of small and medium-sized enterprises, could promote the high-quality development of the edible vegetable oil industry.

In the Shanshenmiaozui (SSMZ) fauna from Nihewan Basin, Canis chihliensis, Nyctereutes sp., Homotherium sp., Acinonyx sp., Lynx shansius and Pachycrocuta licenti have been identified. C. chihliensis is the dominant and best represented species whose fossil materials include crania, mandibles and postcranials, including nearly complete manus and pes. The giant cheetah Acinonyx sp. is only represented by postcranial skeletons (scapula, radius, proximal ulna and an almost complete manus), which represent the richest collection for its kind ever known in China. Nyctereutes sp. is only represented by a partial cranium and distal part of a humerus. Homotherium sp. is only represented by a partial lower m1. Lynx shansius is represented by a nearly complete mandible with p3 and m1 preserved in situ. The fossil materials of Pachycrocuta licenti include a juvenile mandible with dp2–4 and m1 attached, an isolated dp4 and a partial ascending ramus. The SSMZ carnivoran guild resembles those of the classic Nihewan fauna (CNF) and the Dmanisi fauna, which means they probably have a similar geologic age or slightly younger of the SSMZ fauna; and they also share the same open grassland/shrubland habitat. The postcranial bones of C. chihliensis and Acinonyx sp. from SSMZ represent the first records for their kinds in China.

Elevated levels of plasma branched-chain amino acids (BCAAs) have been associated with insulin resistance and type 2 diabetes since the 1960s. Pharmacological activation of branched-chain α-ketoacid dehydrogenase (BCKDH), the rate-limiting enzyme of BCAA oxidation, lowers plasma BCAAs and improves insulin sensitivity. Here we show that modulation of BCKDH in skeletal muscle, but not liver, affects fasting plasma BCAAs in male mice. However, despite lowering BCAAs, increased BCAA oxidation in skeletal muscle does not improve insulin sensitivity. Our data indicate that skeletal muscle controls plasma BCAAs, that lowering fasting plasma BCAAs is insufficient to improve insulin sensitivity and that neither skeletal muscle nor liver account for the improved insulin sensitivity seen with pharmacological activation of BCKDH. These findings suggest potential concerted contributions of multiple tissues in the modulation of BCAA metabolism to alter insulin sensitivity.