arXiv subjects

Rui Zhang

Publications and source records attributed to Rui Zhang.

At least 883 records · Page 49Linked to original sources

arXiv2021

Prefix-to-SQL: Text-to-SQL Generation from Incomplete User Questions

Existing text-to-SQL research only considers complete questions as the input, but lay-users might strive to formulate a complete question. To build a smarter natural language interface to database systems (NLIDB) that also processes incomplete questions, we propose a new task, prefix-to-SQL which takes question prefix from users as the input and predicts the intended SQL. We construct a new benchmark called PAGSAS that contains 124K user question prefixes and the intended SQL for 5 sub-tasks Advising, GeoQuery, Scholar, ATIS, and Spider. Additionally, we propose a new metric SAVE to measure how much effort can be saved by users. Experimental results show that PAGSAS is challenging even for strong baseline models such as T5. As we observe the difficulty of prefix-to-SQL is related to the number of omitted tokens, we incorporate curriculum learning of feeding examples with an increasing number of omitted tokens. This improves scores on various sub-tasks by as much as 9% recall scores on sub-task GeoQuery in PAGSAS.

cs.CL↗

arXiv2021

A General Machine Learning-based Approach for Inverse Design of One-dimensional Photonic Crystals Toward Targeted Visible Light Reflection Spectrum

Data-driven methods have increasingly been applied to the development of optical systems as inexpensive and effective inverse design approaches. Optical properties (e.g., band-gap properties) of photonic crystals (PCs) are closely associated with characteristics of their light reflection spectra. Finding optimal PC constructions (within a pre-specified parameter space) that generate reflection spectra closest to a targeted spectrum is thus an interesting and meaningful inverse design problem, although relevant studies are still limited. Here we report a generally effective machine learning-based inverse design approach for one-dimensional photonic crystals (1DPCs), focusing on visible light spectra which are of high practical relevance. For a given class of 1DPC system, a deep neural network (DNN) in a unified structure is first trained over data from sizeable forward calculations (from layer thicknesses to spectrum). An iterative optimization scheme is then developed based on a coherent integration of DNN backward predictions (from spectrum to layer thicknesses), forward calculations, and Monte Carlo moves. We employ this new approach to four representative 1DPC systems including periodic structures with two-, three-, and four-layer repeating units and a heterostructure. The approach successfully converges to solutions of optimal 1DPC constructions for various targeted spectra regardless of their exact achievability. As two demonstrating examples, inverse designs toward a specially constructed "rectangle-shaped" green-light or red-light reflection spectrum are presented and discussed in detail. Remarkably, the results show that the approach can efficiently find out optimal layer thicknesses even when they are far outside the range covered by the original training data of DNN.

physics.optics↗

arXiv2021

Multi View Spatial-Temporal Model for Travel Time Estimation

Taxi arrival time prediction is essential for building intelligent transportation systems. Traditional prediction methods mainly rely on extracting features from traffic maps, which cannot model complex situations and nonlinear spatial and temporal relationships. Therefore, we propose Multi-View Spatial-Temporal Model (MVSTM) to capture the mutual dependence of spatial-temporal relations and trajectory features. Specifically, we use graph2vec to model the spatial view, dual-channel temporal module to model the trajectory view, and structural embedding to model traffic semantics. Experiments on large-scale taxi trajectory data have shown that our approach is more effective than the existing novel methods. The source code can be found at https://github.com/775269512/SIGSPATIAL-2021-GISCUP-4th-Solution.

cs.LG↗

arXiv2021

TURINGBENCH: A Benchmark Environment for Turing Test in the Age of Neural Text Generation

Recent progress in generative language models has enabled machines to generate astonishingly realistic texts. While there are many legitimate applications of such models, there is also a rising need to distinguish machine-generated texts from human-written ones (e.g., fake news detection). However, to our best knowledge, there is currently no benchmark environment with datasets and tasks to systematically study the so-called "Turing Test" problem for neural text generation methods. In this work, we present the TuringBench benchmark environment, which is comprised of (1) a dataset with 200K human- or machine-generated samples across 20 labels {Human, GPT-1, GPT-2_small, GPT-2_medium, GPT-2_large, GPT-2_xl, GPT-2_PyTorch, GPT-3, GROVER_base, GROVER_large, GROVER_mega, CTRL, XLM, XLNET_base, XLNET_large, FAIR_wmt19, FAIR_wmt20, TRANSFORMER_XL, PPLM_distil, PPLM_gpt2}, (2) two benchmark tasks -- i.e., Turing Test (TT) and Authorship Attribution (AA), and (3) a website with leaderboards. Our preliminary experimental results using TuringBench show that FAIR_wmt20 and GPT-3 are the current winners, among all language models tested, in generating the most human-like indistinguishable texts with the lowest F1 score by five state-of-the-art TT detection models. The TuringBench is available at: https://turingbench.ist.psu.edu/

cs.CL↗

arXiv2021

Wireless Communication Aided by Intelligent Reflecting Surface: Active or Passive?

In this letter, we consider an intelligent reflecting surface (IRS)-aided wireless communication system, where an active or passive IRS is employed to assist the communication between an access point and a user. First, we consider the downlink/uplink communication separately and optimize the IRS placement for rate maximization with an active or passive IRS. We show that the active IRS should be deployed closer to the receiver with the IRS's decreasing amplification power; while in contrast, the passive IRS should be deployed near either the transmitter or receiver. Moreover, with optimized IRS placement, the passive IRS is shown to outperform its active counterpart when the number of reflecting elements is sufficiently large and/or the active-IRS amplification power is too small. Next, we optimize the IRS placement for both active and passive IRSs to maximize the weighted sum-rate of uplink and downlink communications. We show that in this case, the passive IRS is more likely to achieve superior rate performance. This is because the optimal active-IRS placement needs to balance the rate performance in the uplink and downlink, while deploying the passive IRS near the transmitter or receiver is optimal regardless of the uplink or downlink.

cs.IT↗

arXiv2021

EVOQUER: Enhancing Temporal Grounding with Video-Pivoted BackQuery Generation

Temporal grounding aims to predict a time interval of a video clip corresponding to a natural language query input. In this work, we present EVOQUER, a temporal grounding framework incorporating an existing text-to-video grounding model and a video-assisted query generation network. Given a query and an untrimmed video, the temporal grounding model predicts the target interval, and the predicted video clip is fed into a video translation task by generating a simplified version of the input query. EVOQUER forms closed-loop learning by incorporating loss functions from both temporal grounding and query generation serving as feedback. Our experiments on two widely used datasets, Charades-STA and ActivityNet, show that EVOQUER achieves promising improvements by 1.05 and 1.31 at R@0.7. We also discuss how the query generation task could facilitate error analysis by explaining temporal grounding model behavior.

cs.CV↗

arXiv2021

An Exploratory Study on Long Dialogue Summarization: What Works and What's Next

Dialogue summarization helps readers capture salient information from long conversations in meetings, interviews, and TV series. However, real-world dialogues pose a great challenge to current summarization models, as the dialogue length typically exceeds the input limits imposed by recent transformer-based pre-trained models, and the interactive nature of dialogues makes relevant information more context-dependent and sparsely distributed than news articles. In this work, we perform a comprehensive study on long dialogue summarization by investigating three strategies to deal with the lengthy input problem and locate relevant information: (1) extended transformer models such as Longformer, (2) retrieve-then-summarize pipeline models with several dialogue utterance retrieval methods, and (3) hierarchical dialogue encoding models such as HMNet. Our experimental results on three long dialogue datasets (QMSum, MediaSum, SummScreen) show that the retrieve-then-summarize pipeline models yield the best performance. We also demonstrate that the summary quality can be further improved with a stronger retrieval model and pretraining on proper external summarization datasets.

cs.CL↗

arXiv2021

Application of Quantum Machine Learning using the Quantum Kernel Algorithm on High Energy Physics Analysis at the LHC

Quantum machine learning could possibly become a valuable alternative to classical machine learning for applications in High Energy Physics by offering computational speed-ups. In this study, we employ a support vector machine with a quantum kernel estimator (QSVM-Kernel method) to a recent LHC flagship physics analysis: $t\bar{t}H$ (Higgs boson production in association with a top quark pair). In our quantum simulation study using up to 20 qubits and up to 50000 events, the QSVM-Kernel method performs as well as its classical counterparts in three different platforms from Google Tensorflow Quantum, IBM Quantum and Amazon Braket. Additionally, using 15 qubits and 100 events, the application of the QSVM-Kernel method on the IBM superconducting quantum hardware approaches the performance of a noiseless quantum simulator. Our study confirms that the QSVM-Kernel method can use the large dimensionality of the quantum Hilbert space to replace the classical feature space in realistic physics datasets.

quant-ph↗

arXiv2021

Eden: A Unified Environment Framework for Booming Reinforcement Learning Algorithms

With AlphaGo defeats top human players, reinforcement learning(RL) algorithms have gradually become the code-base of building stronger artificial intelligence(AI). The RL algorithm design firstly needs to adapt to the specific environment, so the designed environment guides the rapid and profound development of RL algorithms. However, the existing environments, which can be divided into real world games and customized toy environments, have obvious shortcomings. For real world games, it is designed for human entertainment, and too much difficult for most of RL researchers. For customized toy environments, there is no widely accepted unified evaluation standard for all RL algorithms. Therefore, we introduce the first virtual user-friendly environment framework for RL. In this framework, the environment can be easily configured to realize all kinds of RL tasks in the mainstream research. Then all the mainstream state-of-the-art(SOTA) RL algorithms can be conveniently evaluated and compared. Therefore, our contributions mainly includes the following aspects: 1.single configured environment for all classification of SOTA RL algorithms; 2.combined environment of more than one classification RL algorithms; 3.the evaluation standard for all kinds of RL algorithms. With all these efforts, a possibility for breeding an AI with capability of general competency in a variety of tasks is provided, and maybe it will open up a new chapter for AI.

cs.LG↗

arXiv2021

Cross-Lingual Training with Dense Retrieval for Document Retrieval

Dense retrieval has shown great success in passage ranking in English. However, its effectiveness in document retrieval for non-English languages remains unexplored due to the limitation in training resources. In this work, we explore different transfer techniques for document ranking from English annotations to multiple non-English languages. Our experiments on the test collections in six languages (Chinese, Arabic, French, Hindi, Bengali, Spanish) from diverse language families reveal that zero-shot model-based transfer using mBERT improves the search quality in non-English mono-lingual retrieval. Also, we find that weakly-supervised target language transfer yields competitive performances against the generation-based target language transfer that requires external translators and query generators.

cs.CL↗

arXiv2021

Strange and charm contributions to nucleon charges and moments

We present preliminary results for strange and charm contributions to nucleon charges and moments. The scalar, axial and tensor charges, and unpolarized first moments are calculated using clover-on-HISQ formulation and cover four lattice spacings, $a=\{0.06,0.09,0.12, 0.15\}$~fm, and three pion masses, $M_π=\{310,220,130\}$~MeV. The renormalization factors are calculated nonperturbatively using the RI-sMOM scheme. We carry out a chiral and continuum extrapolation to obtain physical results.

hep-lat↗

arXiv2021

Anchor-Assisted Channel Estimation for Intelligent Reflecting Surface Aided Multiuser Communication

Channel estimation is a practical challenge for intelligent reflecting surface (IRS) aided wireless communication. As the number of IRS reflecting elements or IRS-aided users increases, the channel training overhead becomes excessively high, which results in long delay and low throughput in data transmission. To tackle this challenge, we propose in this paper a new anchor-assisted channel estimation approach, where two anchor nodes, namely A1 and A2, are deployed near the IRS for facilitating its aided base station (BS) in acquiring the cascaded BS-IRS-user channels required for data transmission. Specifically, in the first scheme, the partial channel state information (CSI) on the element-wise channel gain square of the common BS-IRS link for all users is first obtained at the BS via the anchor-assisted training and feedback. Then, by leveraging such partial CSI, the cascaded BS-IRS-user channels are efficiently resolved at the BS with additional training by the users. While in the second scheme, the BS-IRS-A1 and A1-IRS-A2 channels are first estimated via the training by A1. Then, with additional training by A2, all users estimate their individual cascaded A2-IRS-user channels simultaneously. Based on the CSI fed back from A2 and all users, the BS resolves the cascaded BS-IRS-user channels efficiently. In both schemes, the quasi-static channels among the fixed BS, IRS, and two anchors are estimated off-line only, which greatly reduces the real-time training overhead. Simulation results demonstrate that our proposed anchor-assisted channel estimation schemes achieve superior performance as compared to existing IRS channel estimation schemes, under various practical setups. In addition, the first proposed scheme outperforms the second one when the number of antennas at the BS is sufficiently large, and vice versa.

cs.IT↗

arXiv2021

Multi-Beam Multi-Hop Routing for Intelligent Reflecting Surfaces Aided Massive MIMO

Intelligent reflecting surface (IRS) is envisioned to play a significant role in future wireless communication systems as an effective means of reconfiguring the radio signal propagation environment. In this paper, we study a new multi-IRS aided massive multiple-input multiple-output (MIMO) system, where a multi-antenna BS transmits independent messages to a set of remote single-antenna users using orthogonal beams that are subsequently reflected by different groups of IRSs via their respective multi-hop passive beamforming over pairwise line-of-sight (LoS) links. We aim to select optimal IRSs and their beam routing path for each of the users, along with the active/passive beamforming at the BS/IRSs, such that the minimum received signal power among all users is maximized. This problem is particularly difficult to solve due to a new type of path separation constraints for avoiding the IRS-reflected signal induced interference among different users. To tackle this difficulty, we first derive the optimal BS/IRS active/passive beamforming solutions based on their practical codebooks given the reflection paths. Then we show that the resultant multi-beam multi-hop routing problem can be recast as an equivalent graph-optimization problem, which is however NP-complete. To solve this challenging problem, we propose an efficient recursive algorithm to partially enumerate the feasible routing solutions, which is able to effectively balance the performance-complexity trade-off. Numerical results demonstrate that the proposed algorithm achieves near-optimal performance with low complexity and outperforms other benchmark schemes. Useful insights into the optimal multi-beam multi-hop routing design are also drawn under different setups of the multi-IRS aided massive MIMO network.

cs.IT↗

arXiv2021

Distributed Beam Training for Intelligent Reflecting Surface Enabled Multi-Hop Routing

Intelligent reflecting surface (IRS) is an emerging technology to enhance the spectral and energy efficiency of wireless communications cost-effectively. This letter considers a new multi-IRS aided wireless network where a cascaded line-of-sight (LoS) link is established between the base station (BS) and a remote user by leveraging the multi-hop signal reflection of selected IRSs. As compared to the conventional single-/double-hop IRS system, multi-hop IRS system provides more pronounced path diversity and cooperative passive beamforming gains, especially in the environment with dense obstacles. However, a more challenging joint active/passive beamforming and multi-hop beam routing problem also arises for maximizing the end-to-end channel gain. Furthermore, the number of IRS-associated channel coefficients increases drastically with the number of IRS hops. To tackle the above issues, in this letter we propose a new and efficient beam training based solution by considering the use of practical codebook-based BS/IRS active/passive beamforming without the need of explicit channel estimation. Instead of exhaustively or sequentially searching over all combinations of active and passive beam patterns for each beam route, a distributed beam training scheme is proposed to reduce the complexity, by exploiting the (nearly) time-invariant BS-IRS and inter-IRS channels and the cooperative training among the BS and IRSs' controllers. Simulation results show that our proposed design achieves the end-to-end channel gain close to that of the sequential beam search, but at a much lower training overhead and complexity.

cs.IT↗

arXiv2021

Pion Distribution Amplitudes in the Continuum Limit

We present a lattice-QCD calculation of the pion distribution amplitudes using large-momentum effective theory (LaMET). Our calculation is carried out using five ensembles with 2+1+1 flavors of highly improved staggered quarks (HISQ), generated by MILC collaboration, at 310 MeV and 220 MeV pion mass with 0.06, 0.09, 0.12 and 0.15 fm lattice spacings. We use clover fermion action for the valence quarks and tune the quark mass to match the lightest light and strange masses in the sea. The resulting lattice matrix elements are nonperturbatively renormalized in regularization-independent momentum-subtraction (RI/MOM) scheme and extrapolated to the continuum. We compare different approaches to extract the x-dependence of the pion distribution amplitudes.

hep-lat↗

arXiv2021

Hindsight Value Function for Variance Reduction in Stochastic Dynamic Environment

Policy gradient methods are appealing in deep reinforcement learning but suffer from high variance of gradient estimate. To reduce the variance, the state value function is applied commonly. However, the effect of the state value function becomes limited in stochastic dynamic environments, where the unexpected state dynamics and rewards will increase the variance. In this paper, we propose to replace the state value function with a novel hindsight value function, which leverages the information from the future to reduce the variance of the gradient estimate for stochastic dynamic environments. Particularly, to obtain an ideally unbiased gradient estimate, we propose an information-theoretic approach, which optimizes the embeddings of the future to be independent of previous actions. In our experiments, we apply the proposed hindsight value function in stochastic dynamic environments, including discrete-action environments and continuous-action environments. Compared with the standard state value function, the proposed hindsight value function consistently reduces the variance, stabilizes the training, and improves the eventual policy.

cs.LG↗

arXiv2021

Logic-Consistency Text Generation from Semantic Parses

Text generation from semantic parses is to generate textual descriptions for formal representation inputs such as logic forms and SQL queries. This is challenging due to two reasons: (1) the complex and intensive inner logic with the data scarcity constraint, (2) the lack of automatic evaluation metrics for logic consistency. To address these two challenges, this paper first proposes SNOWBALL, a framework for logic consistent text generation from semantic parses that employs an iterative training procedure by recursively augmenting the training set with quality control. Second, we propose a novel automatic metric, BLEC, for evaluating the logical consistency between the semantic parses and generated texts. The experimental results on two benchmark datasets, Logic2Text and Spider, demonstrate the SNOWBALL framework enhances the logic consistency on both BLEC and human evaluation. Furthermore, our statistical analysis reveals that BLEC is more logically consistent with human evaluation than general-purpose automatic metrics including BLEU, ROUGE and, BLEURT. Our data and code are available at https://github.com/Ciaranshu/relogic.

cs.CL↗

arXiv2021

Neat: Low-Complexity, Efficient On-Chip Cache Coherence

Cache coherence protocols such as MESI that use writer-initiated invalidation have high complexity and sometimes have poor performance and energy usage, especially under false sharing. Such protocols require numerous transient states, a shared directory, and support for core-to-core communication, while also suffering under false sharing. An alternative to MESI's writer-initiated invalidation is self-invalidation, which achieves lower complexity than MESI but adds high performance costs or relies on programmer annotations or specific data access patterns. This paper presents Neat, a low-complexity, efficient cache coherence protocol. Neat uses self-invalidation, thus avoiding MESI's transient states, directory, and core-to-core communication requirements. Neat uses novel mechanisms that effectively avoid many unnecessary self-invalidations. An evaluation shows that Neat is simple and has lower verification complexity than the MESI protocol. Neat not only outperforms state-of-the-art self-invalidation protocols, but its performance and energy consumption are comparable to MESI's, and it outperforms MESI under false sharing.

cs.AR↗