Search

Reactions of (η4-benzylideneacetone)Fe(CO)3 and the α,ω-diphosphines Ar2P(CH2)nPAr2 afford the trigonal bipyramidal diiron tetraphosphorus complexes trans,trans-(CO)3Fe[Ar2P(CH2)nPAr2]2Fe(CO)3 (n/Ar = 3/Ph 3, 4/Ph 4a, 4/p-tol 4b; 56-19%). Crystal structures establish essentially parallel P-Fe-P axes, iron-iron distances of 5.894(9)-5.782(1) A (3) and 6.403(1)-6.466(1) A (4a,b), and van der Waals radii of 4.45 A for the Fe(CO)3 rotators, the planes of which are offset by 0.029-1.665 A. Analogous reactions of Ph2P(CH2)6PPh2 yield the square pyramidal monoiron complex trans-(CO)3Fe[Ph2P(CH2)6PPh2] (6', 31%), a rare case where a diphosphine spans trans basal positions (∠P-Fe-P 147.4(2)°). Both 3 and 6' exhibit two CO 13C NMR signals at room temperature, indicating slow exchange on the NMR time scale, which in the former could entail Fe(CO)3/Fe(CO)3 gearing. Under analogous conditions, 4a,b exhibit one signal. Previously reported adducts of Fe(CO)3 and Ph2P(CH2)nPPh2 are surveyed (1:1, n = 1-5; 2:2, n = 5), and the IR νC≡O band patterns and energies of all complexes analyzed with the aid of DFT calculations. The diiron complexes are preferred thermodynamically. Attention is given to limiting types of Fe(CO)3/Fe(CO)3 interactions in the diiron complexes.

Large-scale cross-lingual pre-trained language models (xPLMs) have shown effectiveness in cross-lingual sequence labeling tasks (xSL), such as cross-lingual machine reading comprehension (xMRC) by transferring knowledge from a high-resource language to low-resource languages. Despite the great success, we draw an empirical observation that there is a training objective gap between pre-training and fine-tuning stages: e.g., mask language modeling objective requires local understanding of the masked token and the span-extraction objective requires global understanding and reasoning of the input passage/paragraph and question, leading to the discrepancy between pre-training and xMRC. In this paper, we first design a pre-training task tailored for xSL named Cross-lingual Language Informative Span Masking (CLISM) to eliminate the objective gap in a self-supervised manner. Second, we present ContrAstive-Consistency Regularization (CACR), which utilizes contrastive learning to encourage the consistency between representations of input parallel sequences via unsupervised cross-lingual instance-wise training signals during pre-training. By these means, our methods not only bridge the gap between pretrain-finetune, but also enhance PLMs to better capture the alignment between different languages. Extensive experiments prove that our method achieves clearly superior results on multiple xSL benchmarks with limited pre-training data. Our methods also surpass the previous state-of-the-art methods by a large margin in few-shot data settings, where only a few hundred training examples are available., Comment: 15 pages

As an effective method for high-dimensional information processing, tensor factorization has been widely used in context-aware recommendation. However, the multi-dimensional processing of tensor factorization usually leads to more serious data sparsity. Making full use of the data in recommendation systems, including a small amount of explicit rating data and a large amount of implicit feedback data such as clicks, plays, and purchases, can effectively alleviate this problem. Meanwhile, the processing of sparse explicit data and huge implicit feedback data makes tensor factorization face a severe challenge of computational efficiency. Therefore, this study proposes a fast tensor factorization model named FWCP that unifies explicit and implicit feedback information, which mainly studies as follows: (1) Assigning non-uniform weights to observed and unobserved values to unify explicit and implicit feedback information; (2) Memorizing intermediate data which generated by element-wise alternating least squares(ALS), and implementing parallel optimization of FWCP, which can reduce the training time of tensor factorization based on whole-data. The evaluation results on two real datasets showed that FWCP proposed in this study is superior to other context-aware recommendation algorithms.

Reactions of (η

Lack of training data presents a grand challenge to scaling out spoken language understanding (SLU) to low-resource languages. Although various data augmentation approaches have been proposed to synthesize training data in low-resource target languages, the augmented data sets are often noisy, and thus impede the performance of SLU models. In this paper we focus on mitigating noise in augmented data. We develop a denoising training approach. Multiple models are trained with data produced by various augmented methods. Those models provide supervision signals to each other. The experimental results show that our method outperforms the existing state of the art by 3.05 and 4.24 percentage points on two benchmark datasets, respectively. The code will be made open sourced on github., Comment: Long paper at EMNLP 2021

Producing the embedding of a sentence in an unsupervised way is valuable to natural language matching and retrieval problems in practice. In this work, we conduct a thorough examination of pretrained model based unsupervised sentence embeddings. We study on four pretrained models and conduct massive experiments on seven datasets regarding sentence semantics. We have there main findings. First, averaging all tokens is better than only using [CLS] vector. Second, combining both top andbottom layers is better than only using top layers. Lastly, an easy whitening-based vector normalization strategy with less than 10 lines of code consistently boosts the performance.

Background: Application of platelet-rich plasma (PRP) can improve tendon-bone healing (TBH) after rotator cuff surgical repair. Graphene Oxide (GO) is a steady, controlled, and sustained carrier. The purpose of this study is to determine whether GO/PRP Composite Scaffold enhances the TBH after RC surgical repair in a rabbit model.Methods: A full-thickness tear of the supraspinatus tendon was created and repaired in 36 adult male New Zealand rabbits. They were divided into three groups: Control group, PRP group, and GO/PRP Composite Scaffold group (GO group). The effect of GO/PRP Composite Scaffold on TBH was assessed using histological and biomechanical evaluations at 8 and 12 weeks postoperatively.Results: Histological analysis showed that greater continuity, better orientation, and more density of collagen fiber were detected in the GO group than PRP and Control groups at 8 and 12 weeks, respectively. Results of biomechanical evaluations showed that the load to failure and stiffness of the GO group were statistically higher than those of PRP and Control groups at both 8 and 12 weeks (P0.05), while the stiffness at 12 weeks was higher than that at 8 weeks (P Conclusions: These results demonstrated that GO/PRP Composite Scaffold enhanced the TBH following rotator cuff surgical repair in a rabbit model. The GO may be an effective carrier for PRP into repair sites.

With a rigid fused polyaromatic framework and a well-defined, highly symmetric molecular geometry, pentiptycenes are appealing building blocks for a variety of materials applications. Unfortunately, their use has been limited by the lengthy syntheses of their functionalized derivatives. This communication describes a highly efficient, brief, divergent paradigm for the preparation of octakis-substituted pentiptycene derivatives that starts with the preparation of an octakis(bromo) compound, which can be used as a Pd(0)-catalyzed coupling partner with suitable organometallic compounds to install a range of groups in high yields at the peripheral 2,3,6,7,14,15,19,20 positions, including methyl, allyl, vinyl, alkynyl, aryl, heteroaryl, and even bulky 4-(triphenylmethyl)phenyl substituents.

Numerous valuable efforts have been devoted to achieving arbitrary style transfer since the seminal work of Gatys et al. However, existing state-of-the-art approaches often generate insufficiently stylized results under challenging cases. We believe a fundamental reason is that these approaches try to generate the stylized result in a single shot and hence fail to fully satisfy the constraints on semantic structures in the content images and style patterns in the style images. Inspired by the works on error-correction, instead, we propose a self-correcting model to predict what is wrong with the current stylization and refine it accordingly in an iterative manner. For each refinement, we transit the error features across both the spatial and scale domain and invert the processed features into a residual image, with a network we call Error Transition Network (ETNet). The proposed model improves over the state-of-the-art methods with better semantic structures and more adaptive style pattern details. Various qualitative and quantitative experiments show that the key concept of both progressive strategy and error-correction leads to better results. Code and models are available at https://github.com/zhijieW94/ETNet., Comment: Accepted by NeurIPS 2019

The effect of Rh- and Ir-centers on the optical properties of the BODIPY core has been studied. To this end, novel metal complexes tethered to BODIPY have been prepared through an easy and versatile procedure using N-directed C–H activation reactions. The organometallic moiety has a tremendous influence on the emissive properties of the BODIPY fragment. A photoinduced electron transfer (PET) mechanism is suggested to be the main mechanism responsible for the suppression of the BODIPY fluorescence emission in the newly formed dyads. The efficiency of the PET depends on both the distance between the chromophores in the dyad and the nature of the transition metal (Rh vs. Ir).

With a rigid fused polyaromatic framework and a well-defined, highly symmetric molecular geometry, pentiptycenes are appealing building blocks for a variety of materials applications. Unfortunately, their use has been limited by the lengthy syntheses of their functionalized derivatives. This communication describes a highly efficient, brief, divergent paradigm for the preparation of octakis-substituted pentiptycene derivatives that starts with the preparation of an octakis(bromo) compound, which can be used as a Pd(0)-catalyzed coupling partner with suitable organometallic compounds to install a range of groups in high yields at the peripheral 2,3,6,7,14,15,19,20 positions, including methyl, allyl, vinyl, alkynyl, aryl, heteroaryl, and even bulky 4-(triphenylmethyl)phenyl substituents.

The synthesis, structure and electronic properties of novel Group 6 Fischer alkoxy-bis(carbene) complexes are reported. The UV/Vis spectra of these species display two main absorptions at approximately 350 and 550 nm attributable to a ligand-field (LF) and metal-to-ligand charge-transfer (MLCT) transitions, respectively. The planarity of the system and the cooperative effect of both pentacarbonyl metal moieties greatly enhance the conjugation between the group at the end of the spacer and the metal carbene fragment provoking dramatic changes in the LF and MLCT absorptions. This is in contrast to related push-pull Fischer monocarbenes, where the position of the MLCT band remains mostly unaltered regardless the substituent attached to the donor fragment. In addition, the MLCT maxima can be tuned with subtle modifications of the electronic nature of the central aryl fragment in the novel A-π-D-π-A (A = acceptor, D = donor) systems. DFT and time-dependent (TD) DFT quantum chemical calculations at the B3LYP/def2-SVP level have also been performed to determine the minimum-energy molecular structure of this family of compounds and to analyse the nature of the vertical one-electron excitations associated to the observed UV/Vis absorptions as well as to rationalise their electrochemical behaviour. The ability of tuning up the electronic properties of the compounds studied herein may be of future use in material chemistry.

The systematic investigation of the effect of different catalysts and additives in the reaction of self-dimerization of alkoxychromium(0) (Fischer) carbene complexes resulted in the selection of Pd(P(t)Bu(3))(2) to effect this transformation with good to excellent E selectivities and acceptable to excellent chemical yields. This catalyst will allow the control of the geometry in the synthesis of polyconjugated olefins, one emerging application of these catalytic reactions.

The use of Reclaimed Asphalt Pavement (RAP) material is widespread although the overall amount in the mix and the plant technology to handle it vary for each country. Reusing RAP allows for potential benefits due to less material consumption and possible energy savings although the benefit greatly varies depending upon the technology adopted (hot, warm, or cold mixes) and material performance during the service life. The paper investigates the use of rejuvenator to recover 100% cold RAP mix and evaluates its possible recycle. In particular, the rejuvenated asphalt binder was analysed through several laboratory experimentations including: standard physical tests, rheology tests in the in-service temperature domain, and binder fatigue resistance. The Pressure Aging Vessel (PAV) was also adopted to simulate the in-service aging of the rejuvenated binder. Results highlighted how the rejuvenator provided a differential effect depending on the amount of the rejuvenator which greatly affected the final performance; in particular, a greater content of rejuvenator helps improving performance in the low-temperature domain. Fatigue damage of the binder was increased (compared to oxidized RAP binder without rejuvenator) regardless of the rejuvenator content and the behaviour was similar to virgin bitumen. Finally, the adoption of rejuvenators can be considered as an optimal way to improve the recycling of RAP although the content and type should be carefully calibrated depending upon the RAP and in-service climate conditions of the recycled pavement.

The synthesis, structure, and complete characterization of mono- and bimetallic dyads joining Fischer carbene complexes and BODIPY chromophores are reported. In these organometallic species, the Fischer carbene complex is attached to the BODIPY moiety through a p-aminophenyl group linked at the C8 carbon atom of the BODIPY core. The photophysical properties, namely the corresponding UV/vis absorption and emission spectra of these new metal-carbene complexes, are analyzed and discussed. It is found that whereas the absorption of the considered dyads strongly resembles that of the parent 4-anilinyl-substituted BODIPY, the fluorescence emission is significantly reduced in these species, very likely as a result of a Förster-type energy transfer mechanism. At variance, the replacement of the pentacarbonyl-metal(0) fragment by a carbonyl group leads to high fluorescence emission intensity. In addition, the emissive properties of the BODIPY core in these organometallic dyads can be tuned by remote groups by means of π-conjugation, as supported by density functional theory calculations.

The synthesis, structure, and photophysical properties of novel BODIPY-Fischer alkoxy-, thio-, and aminocarbene dyads are reported. The BODIPY chromophore is directly attached to the carbene ligand by an ethylenic spacer, thus forming donor-bridge-acceptor π-extended systems. The extension of the π-conjugation is decisive in the equilibrium geometries of the dyads and is clearly reflected in the corresponding absorption and emission spectra. Whereas the BODIPY fragment is mainly isolated in aminocarbene complexes, it is fully conjugated in alkoxycarbene derivatives. The former thus exhibit the characteristic photophysical properties of BODIPY units, whereas complete suppression of the BODIPY fluorescence emission is observed in the latter, as a direct consequence of the strong electron-accepting character of the (CO)5M=C moiety. As the π-acceptor character of the metal-carbene group can be modified, the electronic properties of the conjugated BODIPY can be tuned. Density functional calculations have been carried out to gain insight into the photophysical properties.

A direct calculation of the mixed-action parameter $��_{mix}$ with valence overlap fermions on a domain-wall fermion sea is presented. The calculation is performed on four ensembles of the 2+1-flavor domain-wall gauge configurations: $24^3 \times 64$ ($a m_l= 0.005$, $a=0.114\fm$) and $32^3 \times 64$ ($a m_l = 0.004, 0.006, 0.008$, $a=0.085\fm$). For pion masses close to $300\MeV$ we find \hbox{$��_{mix}=0.030(6)\GeV^4$} at $a=0.114\fm$ and $��_{mix}=0.033(12)\GeV^4$ at $a=0.085\fm$. The results are quite independent of the lattice spacing and they are significantly smaller than the results for valence domain-wall fermions on Asqtad sea or those of valence overlap fermions on clover sea. Combining the results extracted from these two ensembles, we get $��_{mix}=0.030(6)(5)\GeV^4$, where the first error is statistical and the second is the systematic error associated with the fitting method., 6 pages, 6 figures

The calculation of the strangeness and charmness of the nucleon is presented with overlap fermion action on 2+1 flavor domain wall fermion configurations. We adopt stochastic grid sources and the low mode substitution technique to improve the signals of nucleon correlation functions and the loops. The calculation is done on a $24^3\times 64$ lattice with $m_l=0.005$, $m_h=0.04$, and $a^{-1}=1.73\,{\rm GeV}$. We find $ f_{T_{s}} = 0.048(15)$ and $f_{T_{c}} = 0.029(43)$., Comment: Talk presented at the XXIX International Symposium on Lattice Field Theory, Squaw Valley, Lake Tahoe, California, USA, July 10-16, 2011. 7 pages, 5 figures

We have investigated the electrically active deep level defects in n-type 6H silicon carbide through the use of a series of complimentary spectroscopic techniques such as deep level transient spectroscopy, positron annihilation spectroscopy and photolumniescence. The deep level defects were created by neutron irrdiation, He implantation and electron irradiation with different energies. After analysis of the information gained from the different types of spectroscopy, as well as consideration of the defect creation and annealing behavior under different controlled environments, we provide experimental evidence for the microstructure of certain important deep level defects.

Malignant degeneration is an unusual complication of chronic primary or recurrent pilonidal disease. To date, only 43 cases have been reported, the great majority being squamous cell type. Historically, these cases were treated with primary excision. Recurrence following such treatment was high and prognosis poor compared with nonmelanoma skin cancer. We describe a patient with squamous cell carcinoma arising in chronic pilonidal disease, who is disease-free 3 years after adjuvant radiation therapy following radical excision. This communication describes the clinicopathologic features of 38 cases of squamous cell carcinoma arising in pilonidal disease and suggests treatment modalities.

In this paper, we introduce XGLUE, a new benchmark dataset that can be used to train large-scale cross-lingual pre-trained models using multilingual and bilingual corpora and evaluate their performance across a diverse set of cross-lingual tasks. Comparing to GLUE(Wang et al., 2019), which is labeled in English for natural language understanding tasks only, XGLUE has two main advantages: (1) it provides 11 diversified tasks that cover both natural language understanding and generation scenarios; (2) for each task, it provides labeled data in multiple languages. We extend a recent cross-lingual pre-trained model Unicoder(Huang et al., 2019) to cover both understanding and generation tasks, which is evaluated on XGLUE as a strong baseline. We also evaluate the base versions (12-layer) of Multilingual BERT, XLM and XLM-R for comparison.

Recent research demonstrates the effectiveness of using pretrained language models (PLM) to improve dense retrieval and multilingual dense retrieval. In this work, we present a simple but effective monolingual pretraining task called contrastive context prediction (CCP) to learn sentence representation by modeling sentence level contextual relation. By pushing the embedding of sentences in a local context closer and pushing random negative samples away, different languages could form isomorphic structure, then sentence pairs in two different languages will be automatically aligned. Our experiments show that model collapse and information leakage are very easy to happen during contrastive training of language model, but language-specific memory bank and asymmetric batch normalization operation play an essential role in preventing collapsing and information leakage, respectively. Besides, a post-processing for sentence embedding is also very effective to achieve better retrieval performance. On the multilingual sentence retrieval task Tatoeba, our model achieves new SOTA results among methods without using bilingual data. Our model also shows larger gain on Tatoeba when transferring between non-English pairs. On two multi-lingual query-passage retrieval tasks, XOR Retrieve and Mr.TYDI, our model even achieves two SOTA results in both zero-shot and supervised setting among all pretraining models using bilingual data.

Supplemental Digital Content is available in the text., Background: The novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the cause of a rapidly spreading illness, Coronavirus Disease 2019 (COVID-19), affecting thousands of people around the world. Urgent guidance for clinicians caring for the sickest of these patients is needed. Methods: We formed a panel of 36 experts from 12 countries. All panel members completed the World Health Organization conflict of interest disclosure form. The panel proposed 53 questions that are relevant to the management of COVID-19 in the ICU. We searched the literature for direct and indirect evidence on the management of COVID-19 in critically ill patients in the ICU. We identified relevant and recent systematic reviews on most questions relating to supportive care. We assessed the certainty in the evidence using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach, then generated recommendations based on the balance between benefit and harm, resource and cost implications, equity, and feasibility. Recommendations were either strong or weak, or in the form of best practice recommendations. Results: The Surviving Sepsis Campaign COVID-19 panel issued 54 statements, of which four are best practice statements, nine are strong recommendations, and 35 are weak recommendations. No recommendation was provided for six questions. The topics were: 1) infection control, 2) laboratory diagnosis and specimens, 3) hemodynamic support, 4) ventilatory support, and 5) COVID-19 therapy. Conclusion: The Surviving Sepsis Campaign COVID-19 panel issued several recommendations to help support healthcare workers caring for critically ill ICU patients with COVID-19. When available, we will provide new evidence in further releases of these guidelines.

Here we introduce a web mining approach for automatically identifying new product ideas extracted from web logs. A web log - also known as blog - is a web site that provides commentary, news, and further information on a subject written by individual persons. We can find a large amount of web logs for nearly each topic where consumers present their needs for new products. These new product ideas probably are valuable for producers as well as for researchers and developers. This is because they can lead to a new product development process. Finding these new product ideas is a well-known task in marketing. Therefore, with this automatic approach we support marketing activities by extracting new and useful product ideas from textual information in internet logs. This approach is implemented by a web-based application named Product Idea Web Log Miner where users from the marketing department provide descriptions of existing products. As a result, new product ideas are extracted from the web logs and presented to the users.