Large language models have recently been shown to attain reasonable zero-shot generalization on a diverse set of tasks (Brown et al., 2020). It has been hypothesized that this is a consequence of implicit multitask learning in language model training . Can zero-shot generalization instead be directly induced by explicit multitask learning? To test this question at scale, we develop a system for easily mapping general natural language tasks into a human-readable prompted form. We convert a large set of supervised datasets, each with multiple prompts using varying natural language. These prompted datasets allow for benchmarking the ability of a model to perform completely unseen tasks specified in natural language. We fine-tune a pretrained encoder-decoder model on this multitask mixture covering a wide variety of tasks. The model attains strong zero-shot performance on several standard datasets, often outperforming models up to 16× its size. Further, our approach attains strong performance on a subset of tasks from the BIG-Bench benchmark, outperforming models up to 6× its size. All prompts and trained models are available at github.com/bigscience-workshop/promptsource/ and huggingface.co/bigscience/T0pp.

The role of the Quaternary ice ages in forming the contemporary genetic structure of populations has been well studied in a number of global regions. However, due to the different nature of glaciations and complex topography, their role in shaping eastern Eurasian genetic diversity, particular in areas surrounding the Tibetan Plateau have remained largely unstudied. We aimed to address this question by examining the genetic structure of an alpine forest-associated taxon, the blood pheasant (Ithaginis cruentus) to infer its phylogeographic history. We detected three phylogenetic lineages and four current population groups. By comparing molecular and palaeovegetation data, we found that major glaciations during the Pleistocene have had a major impact upon the current genetic diversity of this species. Coalescent simulations indicate that the populations retreated to different refugia during some glacial periods in the Pleistocene, but persisted through the last glacial maximum (LGM). The most significant recent population expansion was found to have occurred before the LGM, during which palaeoclimatic data indicate that the climate was both warmer and wetter than today. In contrast, during the LGM populations may have adopted an altitudinal shift strategy in order to track changes in alpine glaciers, exemplifying a general response for montane species in the region where alpine glaciations were not large enough to cause qualitative changes in vegetation. Although analysis based on a plumage related gene showed that divergent selection may have contributed to current patterns of intra-specific diversity, demographic isolation is inferred to have played a more dominant role.

An improved approach to assemble ultrathin and thickness-tunable polypyrrole (PPy) films onto multiwall carbon nanotubes (MWCNTs) has been investigated. A facile procedure is demonstrated for controlling the morphology and thickness of PPy film by adding ethanol in the reaction system and a possible mechanism of the coating formation process is proposed. The coated PPy films can be easily tuned by adding ethanol and adjusting a mass ratio of pyrrole to MWCNTs. Moreover, the thickness of PPy significantly influences the electronic conductivity and capacitive behavior of the PPy/MWCNT composites. The method may provide a facile strategy for tailoring the polymer coating on carbon nanotubes (CNTs) for carbon-based device applications.

Room temperature hydrophilic ionic liquids, 1-methyl-3-ethylimidazolium salts containing ethyl-sulfate anions, have dramatic effects on the morphology changes of electrochemically grown Cu 2 O crystals at room temperature in aqueous solutions. The shape of Cu 2 O crystals evolves from cubic to octahedral and spherical shape only by adding a varied small amount of the ionic liquids in the deposited solutions. The possible mechanism has been explored, and the ethyl-sulfate anion is believed to play a key role in the morphology control.

BackgroundInvertases (INVs) are key enzymes regulating sucrose metabolism and are here revealed to be involved in responses to environmental stress in plants. To date, individual members of the invertase gene family and their expression patterns are unknown in sugarcane due to its complex genome despite their significance in sucrose metabolism.ResultsIn this study, based on comparative genomics, eleven cDNA and twelve DNA sequences belonging to 14 non-redundant members of the invertase gene family were successfully cloned from sugarcane. A comprehensive analysis of the invertase gene family was carried out, including gene structures, phylogenetic relationships, functional domains, conserved motifs of proteins. The results revealed that the 14 invertase members from sugarcane could be clustered into three subfamilies, including 6 neutral/alkaline invertases (ShN/AINVs), and 8 acid invertases (ShAINVs). Faster divergence occurred in acid INVs than in neutral/alkaline INVs after the split of sugarcane and sorghum. At least a one-time gene duplication event was observed to have occurred in the four groups of acid INVs, whereas ShN/AINV1 and ShN/AINV2 in the β8 lineage were revealed to be the most recently duplicated genes among their paralogous genes in the β group of N/AINVs. Furthermore, comprehensive expression analysis of these genes was performed in sugarcane seedlings subjected to five abiotic stresses (drought, low temperature, glucose, fructose, and sucrose) using Quantitative Real-time PCR. The results suggested a functional divergence of INVs and their potential role in response to the five different treatments. Enzymatic activity in sugarcane seedlings was detected under five abiotic stresses treatments, and showed that the activities of all INVs were significantly inhibited in response to five different abiotic stresses, and that the neutral/alkaline INVs played a more prominent role in abiotic stresses than the acid INVs.ConclusionsIn this study, we determined the INV gene family members of sugarcane by PCR cloning using sorghum as a reference, providing the first study of the INV gene family in sugarcane. Combining existing INV gene data from 7 plants with a comparative approach including a series of comprehensive analyses to isolate and identify INV gene family members proved to be highly successful. Moreover, the expression levels of INV genes and the variation of enzymatic activities associated with drought, low temperature, glucose, fructose, and sucrose are reported in sugarcane for the first time. The results offered useful foundation and framework for future research for understanding the physiological roles of INVs for sucrose accumulation in sugarcane.Electronic supplementary materialThe online version of this article (doi:10.1186/s12870-017-1052-0) contains supplementary material, which is available to authorized users.

No abstract

No abstract

No abstract