Over the last few years, contextualized pre-trained transformer models such as BERT have provided substantial improvements on information retrieval tasks. Traditional sparse retrieval methods such as BM25 rely on high-dimensional, sparse, bag-of-words query representations to retrieve documents. On the other hand, recent approaches based on pre-trained transformer models such as BERT, fine-tune dense low-dimensional contextualized representations of queries and documents in embedding space. While these dense retrievers enjoy substantial retrieval effectiveness improvements compared to sparse retrievers, they are computationally intensive, requiring substantial GPU resources, and dense retrievers are known to be more expensive from both time and resource perspectives. In addition, sparse retrievers have been shown to retrieve complementary information with respect to dense retrievers, leading to proposals for hybrid retrievers. These hybrid retrievers leverage low-cost, exact-matching based sparse retrievers along with dense retrievers to bridge the semantic gaps between query and documents. In this work, we address this trade-off between the cost and utility of sparse vs dense retrievers by proposing a classifier to select a suitable retrieval strategy (i.e., sparse vs. dense vs. hybrid) for individual queries. Leveraging sparse retrievers for queries which can be answered with sparse retrievers decreases the number of calls to GPUs. Consequently, while utility is maintained, query latency decreases. Although we use less computational resources and spend less time, we still achieve improved performance. Our classifier can select between sparse and dense retrieval strategies based on the query alone. We conduct experiments on the MS MARCO passage dataset demonstrating an improved range of efficiency/effectiveness trade-offs between purely sparse, purely dense or hybrid retrieval strategies, allowing an appropriate strategy to be selected based on a target latency and resource budget.
Over the last few years, contextualized pre-trained transformer models such as BERT have provided substantial improvements on information retrieval tasks. Recent approaches based on pre-trained transformer models such as BERT, fine-tune dense low-dimensional contextualized representations of queries and documents in embedding space. While these dense retrievers enjoy substantial retrieval effectiveness improvements compared to sparse retrievers, they are computationally intensive, requiring substantial GPU resources, and dense retrievers are known to be more expensive from both time and resource perspectives. In addition, sparse retrievers have been shown to retrieve complementary information with respect to dense retrievers, leading to proposals for hybrid retrievers. These hybrid retrievers leverage low-cost, exact-matching based sparse retrievers along with dense retrievers to bridge the semantic gaps between query and documents. In this work, we address this trade-off between the cost and utility of sparse vs dense retrievers by proposing a classifier to select a suitable retrieval strategy (i.e., sparse vs. dense vs. hybrid) for individual queries. Leveraging sparse retrievers for queries which can be answered with sparse retrievers decreases the number of calls to GPUs. Consequently, while utility is maintained, query latency decreases. Although we use less computational resources and spend less time, we still achieve improved performance. Our classifier can select between sparse and dense retrieval strategies based on the query alone. We conduct experiments on the MS MARCO passage dataset demonstrating an improved range of efficiency/effectiveness trade-offs between purely sparse, purely dense or hybrid retrieval strategies, allowing an appropriate strategy to be selected based on a target latency and resource budget.
There has become of increasing interest in transcranial alternating current stimulation (tACS) since its inception nearly a decade ago. tACS in modulating brain state is an active area of research and has been demonstrated effective in various neuropsychological and clinical domains. In the visual domain, much effort has been dedicated to brain rhythms and rhythmic stimulation, i.e., tACS. However, little is known about the interplay between the rhythmic stimulation and visual stimulation. Here, we used steady-state visual evoked potential (SSVEP), induced by flickering driving as a widely used technique for frequency-tagging, to investigate the aftereffect of tACS in healthy human subjects. Seven blocks of 64-channel electroencephalogram were recorded before and after the administration of 20-min 10-Hz tACS, while subjects performed several blocks of SSVEP tasks. We characterized the physiological properties of tACS aftereffect by comparing and validating the temporal, spatial, spatiotemporal and signal-to-noise ratio (SNR) patterns between and within blocks in real tACS and sham tACS. Our result revealed that tACS boosted the 10-Hz SSVEP significantly. Besides, the aftereffect on SSVEP was mitigated with time and lasted up to 5 min. Our results demonstrate the feasibility of facilitating the flickering driving by external rhythmic stimulation and open a new possibility to alter the brain state in a direction by noninvasive transcranial brain stimulation.
The 21st century is the one for comprehensive development and utilization of marine resources. Tianjin,as one of the cities which focus on the development of marine economy of the area surrounding the Bohai Sea,needs to select the leading industry based on the status of each industry and the development prospects of marine economy so as to better optimize the industrial structure and develop the marine economy of Tianjin. The paper first analyzed the current situation of marine industries of Tianjin,and according to the standard of leading industry,chose the selection standard which is suitable for the marine economy of Tianjin. Having established the evaluation system for the leading industry selection,we analyzed each industry of marine economy with McKinsey matrix and then chose the leading industries.
Although autosomal-dominant inheritance is believed an important cause of familial clustering Alzheimer's disease (FAD), it covers only a small proportion of FAD incidence, and so we investigated epigenetic memory as an alternative mechanism to contribute for intergenerational AD pathogenesis. Our data in vivo showed that mys-2 of Caenorhabditis elegans that encodes a putative MYST acetyltransferase responsible for H4K16 acetylation modulated AD occurrence. The phenotypic improvements in the parent generation caused by mys-2 disfunction were passed to their progeny due to epigenetic memory, which resulted in similar H4K16ac levels among the candidate target genes of MYS-2 and similar gene expression patterns of the AD-related pathways. Furthermore, the ROS/CDK-5/ATM pathway functioned as an upstream activator of MYS-2. Our study indicated that MYS-2/MOF could be inherited intergenerationally via epigenetic mechanisms in C. elegans and mammalian cell of AD model, providing a new insight into our understanding of the etiology and inheritance of FAD.
Abstract Background Sirtuin 3 (Sirt3) is a controversial regulator of carcinogenesis. It residents in the mitochondria and gradually decays during aging. In this study, we tried to investigate the role of Sirt3 in carcinogenesis and to explore its involvement in metabolic alteration. Methods We generated conditional intestinal epithelium Sirt3-knockout mice by crossing Apc Min/ + ; Villin-Cre with Sirt3 fl/fl (AVS) mice. The deacetylation site of Lon protease-1 (LONP1) was identified with Mass spectrometry. The metabolic flux phenotype was determined by Seahorse bioanalyzer. Results We found that intestinal epithelial cell-specific ablation of Sirt3 promotes primary tumor growth via stabilizing mitochondrial LONP1. Notably, we newly identified that Sirt3 deacetylates human oncogene LONP1 at N terminal residue lysine 145 (K145). The LONP1 hyperacetylation-mutant K145Q enhances oxidative phosphorylation to accelerate tumor growth, whereas the deacetylation-mutant K145R produces calorie-restriction like phenotype to restrain tumorigenesis. Sirt3 deacetylates LONP1 at K145 and subsequently facilitates the ESCRT0 complex sorting and K63-ubiquitination that resulted in the degradation of LONP1. Our results sustain the notion that Sirt3 is a tumor-suppressor to maintain the appropriate ubiquitination and degradation of oncogene LONP1. Conclusion Sirt3 represents a targetable metabolic checkpoint of oncogenesis, which produces energy restriction effects via maintaining LONP1 K145 deacetylation and subsequent K63 ubiquitination.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)