arXiv subjects

Jian Liu

Publications and source records attributed to Jian Liu.

At least 379 records · Page 21Linked to original sources

arXiv2020

Natural Language Inference in Context -- Investigating Contextual Reasoning over Long Texts

Natural language inference (NLI) is a fundamental NLP task, investigating the entailment relationship between two texts. Popular NLI datasets present the task at sentence-level. While adequate for testing semantic representations, they fall short for testing contextual reasoning over long texts, which is a natural part of the human inference process. We introduce ConTRoL, a new dataset for ConTextual Reasoning over Long texts. Consisting of 8,325 expert-designed "context-hypothesis" pairs with gold labels, ConTRoL is a passage-level NLI dataset with a focus on complex contextual reasoning types such as logical reasoning. It is derived from competitive selection and recruitment test (verbal reasoning test) for police recruitment, with expert level quality. Compared with previous NLI benchmarks, the materials in ConTRoL are much more challenging, involving a range of reasoning types. Empirical results show that state-of-the-art language models perform by far worse than educated humans. Our dataset can also serve as a testing-set for downstream tasks like Checking Factual Correctness of Summaries.

cs.CL↗

arXiv2020

Linear-Model-inspired Neural Network for Electromagnetic Inverse Scattering

Electromagnetic inverse scattering problems (ISPs) aim to retrieve permittivities of dielectric scatterers from the scattering measurement. It is often highly nonlinear, caus-ing the problem to be very difficult to solve. To alleviate the issue, this letter exploits a linear model-based network (LMN) learning strategy, which benefits from both model complexity and data learning. By introducing a linear model for ISPs, a new model with network-driven regular-izer is proposed. For attaining efficient end-to-end learning, the network architecture and hyper-parameter estimation are presented. Experimental results validate its superiority to some state-of-the-arts.

eess.IV↗

arXiv2020

Motion Sensor-based Privacy Attack on Smartphones

In this paper, we build a speech privacy attack that exploits speech reverberations generated from a smartphone's in-built loudspeaker captured via a zero-permission motion sensor (accelerometer). We design our attack Spearphone2, and demonstrate that speech reverberations from inbuilt loudspeakers, at an appropriate loudness, can impact the accelerometer, leaking sensitive information about the speech. In particular, we show that by exploiting the affected accelerometer readings and carefully selecting feature sets along with off-the-shelf machine learning techniques, Spearphone can successfully perform gender classification (accuracy over 90%) and speaker identification (accuracy over 80%) for any audio/video playback on the smartphone. Our results with testing the attack on a voice call and voice assistant response were also encouraging, showcasing the impact of the proposed attack. In addition, we perform speech recognition and speech reconstruction to extract more information about the eavesdropped speech to an extent. Our work brings to light a fundamental design vulnerability in many currently-deployed smartphones, which may put people's speech privacy at risk while using the smartphone in the loudspeaker mode during phone calls, media playback or voice assistant interactions.

cs.CR↗

arXiv2020

Rational homotopy type of mapping spaces via cohomology algebras

In this paper, we show that for finite $CW$-complexes $X$ and two-stage space $Y$ (for example $n$-spheres $S^n$, homogeneous spaces and $F_0$-spaces), the rational homotopy type of $\map(X, Y)$ is determined by the cohomology algebra $H^*(X; \Q)$ and the rational homotopy type of $Y$. From this, we deduce the existence of H-structures on a component of the mapping space $\map(X, Y)$, assuming the cohomology algebras of $X$ and $Y$ are isomorphism. Finally, we will show that $\map(X, Y; f)\simeq\map(X, Y; f')$ if the corresponding \emph{Maurer-Cartan elements} are connected by an algebra automorphism of $H^\ast(X, \Q)$.

math.AT↗

arXiv2020

In Situ characterization of the proton enhanced conductivity of 20nm TiO2 thin-films obtained on the surface of optical fiber

We report in situ characterized TiO2 thin-films deposited on optical fiber, having thicknesses in the 20-100nm range, and having enhanced conductivity values of 700S/cm upon interacting with hydrogen. This conductivity was achieved in pure hydrogen at 800-900C, having a measured activation energy of 0.26eV of the hopping type. Given the variability in the observed results, it is postulated that the highest conductivity achievable may be much greater than what is currently demonstrated. The conductivity is retained after cooling to ambient temperatures as confirmed by Hall measurements, and subsequent grazing-incidence x-ray diffraction and TEM measurements show the films to be in the rutile phase. The exceptional conductivity in these films is hypothesized to result from direct proton incorporation into the lattice populating the conduction band with excess electrons, or from altering the Titania lattice to form conductive Magneli phases. The films did not display any evidence of transformations, however formation of Magneli phases was confirmed for powders. These interesting results, observed by examining 20nm films on the surface of optical fiber in combination with the first impedance spectroscopy performed on films on optical fiber in high temperature Fuel Cell type environments, confirm hypotheses arrived at in prior publications where thin-films of Titania had optical properties which could only be explained by the current claim. Titania thin-films on optical-fiber are being explored for high temperature hydrogen derived energy generation, thermo-photonic energy conversion, and associated sensors due to their unique interactions with hydrogen.

cond-mat.mtrl-sci↗

arXiv2020

Singular equivalences induced by bimodules and quadratic monomial algebras

We investigate the problem when the tensor functor by a bimodule yields a singular equivalence. It turns out that this problem is equivalent to the one when the Hom functor given by the same bimodule induces a triangle equivalence between the homotopy categories of acyclic complexes of injective modules. We give conditions on when a bimodule appears in a pair of bimodules, that defines a singular equivalence with level. We construct an explicit bimodule, which yields a singular equivalence between a quadratic monomial algebra and its associated algebra with radical square zero. Under certain conditions which include the Gorenstein cases, the bimodule does appear in a pair of bimodules defining a singular equivalence with level.

math.RT↗

arXiv2020

Comprehensive Control of Metamagnetic Transition of Antiferromagnetic Mott Insulator Sr2IrO4 by in-situ Anisotropic Strain

Metamagnetism in antiferromagnets exhibits distinct critical behaviors and dynamics when invoking spin reversal and rotation. Here we show a 0.05% anisotropic strain suffices to in-situ modulate the metamagnetic critical field of the Mott insulator Sr2IrO4 by over 50%, enabling electrical switching of the transition. Resonant x-ray scattering and model simulation reveal that the transition is completely tuned from the spin-flop to spin-flip type as the strain introduces C4-symmetry-breaking magnetic anisotropy. Simultaneous transport study indicates the metamagnetic responses are reflected in the large elasto- and magnetoconductance, highlighting the active charge degree of freedom in the spin-orbit-coupled Mott state and its potential for spin-electronics.

cond-mat.str-el↗

arXiv2020

LogiQA: A Challenge Dataset for Machine Reading Comprehension with Logical Reasoning

Machine reading is a fundamental task for testing the capability of natural language understanding, which is closely related to human cognition in many aspects. With the rising of deep learning techniques, algorithmic models rival human performances on simple QA, and thus increasingly challenging machine reading datasets have been proposed. Though various challenges such as evidence integration and commonsense knowledge have been integrated, one of the fundamental capabilities in human reading, namely logical reasoning, is not fully investigated. We build a comprehensive dataset, named LogiQA, which is sourced from expert-written questions for testing human Logical reasoning. It consists of 8,678 QA instances, covering multiple types of deductive reasoning. Results show that state-of-the-art neural models perform by far worse than human ceiling. Our dataset can also serve as a benchmark for reinvestigating logical AI under the deep learning NLP setting. The dataset is freely available at https://github.com/lgw863/LogiQA-dataset

cs.CL↗

arXiv2020

Adversarial Training Based Multi-Source Unsupervised Domain Adaptation for Sentiment Analysis

Multi-source unsupervised domain adaptation (MS-UDA) for sentiment analysis (SA) aims to leverage useful information in multiple source domains to help do SA in an unlabeled target domain that has no supervised information. Existing algorithms of MS-UDA either only exploit the shared features, i.e., the domain-invariant information, or based on some weak assumption in NLP, e.g., smoothness assumption. To avoid these problems, we propose two transfer learning frameworks based on the multi-source domain adaptation methodology for SA by combining the source hypotheses to derive a good target hypothesis. The key feature of the first framework is a novel Weighting Scheme based Unsupervised Domain Adaptation framework (WS-UDA), which combine the source classifiers to acquire pseudo labels for target instances directly. While the second framework is a Two-Stage Training based Unsupervised Domain Adaptation framework (2ST-UDA), which further exploits these pseudo labels to train a target private extractor. Importantly, the weights assigned to each source classifier are based on the relations between target instances and source domains, which measured by a discriminator through the adversarial training. Furthermore, through the same discriminator, we also fulfill the separation of shared features and private features. Experimental results on two SA datasets demonstrate the promising performance of our frameworks, which outperforms unsupervised state-of-the-art competitors.

cs.CL↗

arXiv2020

Duality and symmetry of complexity over complete intersections via exterior homology

We study homological properties of a locally complete intersection ring by importing facts from homological algebra over exterior algebras. One application is showing that the thick subcategories of the bounded derived category of a locally complete intersection ring are self-dual under Grothendieck duality. This was proved by Stevenson when the ring is a quotient of a regular ring modulo a regular sequence; we offer two independent proofs in the more general setting. Second, we use these techniques to supply new proofs that complete intersections possess symmetry of complexity.

math.AC↗

arXiv2020

Evolution of clustering structure through the momentum distributions in $^{8-10}$Be isotopes

We investigate the evolution of clustering structure through the momentum distributions in the $^{8-10}$Be isotopes. The nucleon dynamics within the inter-cluster antisymmetrization are discussed via the momentum distribution of a Brink type $α$-$α$ wave function. For the state with a small $α$-$α$ distance, we observe a significant depression with a dip structure at zero-momentum and an enhanced tail at relatively higher momentum region. In addition, we find the "cluster structure" in the intrinsic frame of momentum space, which is complementary to its significant $α$-cluster dissolution in the coordinate space because of the strong antisymmetrization. For the physical $^{8-10}$Be isotopes, the Tohsaki-Horiuchi-Schuck-R{ö}pke (THSR) wave functions are adopted. The evolution from the dilute clustering state to the compact one is demonstrated by a successive depression at the zero-momentum of nucleon distribution for the two $α$-clusters within $^{8-10}$Be isotopes. For the compact $^{10}$Be nucleus, the momentum distribution of all nucleons shows significant depression at zero-momentum with a dip structure, which is found to be contributed by both the inter-cluster antisymmetrization and the $p$-orbit occupation of the valence neutrons. This study proposes a new window for the investigations of the $α$-clustering effects via the low-momentum components of nuclei, which is expected to be extended to the heavier nuclear clustering states.

nucl-th↗

arXiv2020

Real-time, Universal, and Robust Adversarial Attacks Against Speaker Recognition Systems

As the popularity of voice user interface (VUI) exploded in recent years, speaker recognition system has emerged as an important medium of identifying a speaker in many security-required applications and services. In this paper, we propose the first real-time, universal, and robust adversarial attack against the state-of-the-art deep neural network (DNN) based speaker recognition system. Through adding an audio-agnostic universal perturbation on arbitrary enrolled speaker's voice input, the DNN-based speaker recognition system would identify the speaker as any target (i.e., adversary-desired) speaker label. In addition, we improve the robustness of our attack by modeling the sound distortions caused by the physical over-the-air propagation through estimating room impulse response (RIR). Experiment using a public dataset of 109 English speakers demonstrates the effectiveness and robustness of our proposed attack with a high attack success rate of over 90%. The attack launching time also achieves a 100X speedup over contemporary non-universal attacks.

eess.AS↗

arXiv2020

Strain-modulated Slater-Mott crossover of pseudospin-half square-lattice in (SrIrO3)1/ (SrTiO3)1 superlattices

We report on the epitaxial strain-driven electronic and antiferromagnetic modulations of a pseudospin-half square lattice realized in superlattices of (SrIrO3)1/(SrTiO3)1. With increasing compressive strain, we find the low-temperature insulating behavior to be strongly suppressed with a corresponding systematic reduction of both the Neel temperature and the staggered moment. However, despite such a suppression, the system remains weakly insulating above the Neel transition. The emergence of metallicity is observed under large compressive strain but only at temperatures far above the Néel transition. These behaviors are characteristics of the Slater-Mott crossover regime, providing a unique experimental model system of the spin-half Hubbard Hamiltonian with a tunable intermediate coupling strength.

cond-mat.str-el↗

arXiv2020

Anomalous Magnetoresistance in Centrosymmetric Skyrmion-Lattice Magnet Gd2PdSi3

We performed a systematic study of the temperature- and field-dependence of magnetization and resistivity of Gd2PdSi3, which is a centrosymmetric skyrmion crystal. While the magnetization behavior is consistent with the reported phase diagram based on susceptibility, we show that a phase diagram can also be constructed based on the anomalous magnetoresistance with one-to-one correspondence among all the features. In addition, the crossover boundary into the field-induced ferromagnetic state is also identified. Our results suggest that the ferromagnetic spin fluctuations above the Néel temperature play a key role in the high sensitivity of the resistivity anomalies to magnetic field, pointing to the rich interplay of different magnetic correlations at zero and finite wave vectors underlying the skyrmion lattice in this frustrated itinerant magnet.

cond-mat.str-el↗

arXiv2020

Gate-tunable van der Waals heterostructure for reconfigurable neural network vision sensor

Early processing of visual information takes place in the human retina. Mimicking neurobiological structures and functionalities of the retina provide a promising pathway to achieving vision sensor with highly efficient image processing. Here, we demonstrate a prototype vision sensor that operates via the gate-tunable positive and negative photoresponses of the van der Waals (vdW) vertical heterostructures. The sensor emulates not only the neurobiological functionalities of bipolar cells and photoreceptors but also the unique synaptic connectivity between bipolar cells and photoreceptors. By tuning gate voltage for each pixel, we achieve reconfigurable vision sensor for simultaneously image sensing and processing. Furthermore, our prototype vision sensor itself can be trained to classify the input images, via updating the gate voltages applied individually to each pixel in the sensor. Our work indicates that vdW vertical heterostructures offer a promising platform for the development of neural network vision sensor.

cond-mat.mtrl-sci↗

arXiv2020

Gravitational waves detection with exceptional points in micro cavities

Here we propose a new gravitational waves(GWs) detector in broad frequency band, which is operated at exceptional points(EPs) in micro cavities. The detected signal is an eigenfrequency split of the mechanical modes caused by the spatial strain. Due to the complex square root topology near the EP, the splitting is greatly enhanced for sufficiently small perturbations. Compared to current strategies, it can be achieved at the room temperature and has advantages in micro device scale, wide frequency band and higher sensitivity.

quant-ph↗

arXiv2020

Review and Examination of Input Feature Preparation Methods and Machine Learning Models for Turbulence Modeling

Model extrapolation to unseen flow is one of the biggest challenges facing data-driven turbulence modeling, especially for models with high dimensional inputs that involve many flow features. In this study we review previous efforts on data-driven Reynolds-Averaged Naiver Stokes (RANS) turbulence modeling and model extrapolation, with main focus on the popular methods being used in the field of transfer learning. Several potential metrics to measure the dissimilarity between training flows and testing flows are examined. Different Machine Learning (ML) models are compared to understand how the capacity or complexity of the model affects its behavior in the face of dataset shift. Data preprocessing schemes which are robust to covariate shift, like normalization, transformation, and importance re-weighted likelihood, are studied to understand whether it is possible to find projections of the data that attenuate the differences in the training and test distributions while preserving predictability. Three metrics are proposed to assess the dissimilarity between training/testing dataset. To attenuate the dissimilarity, a distribution matching framework is used to align the statistics of the distributions. These modifications also allow the regression tasks to have better accuracy in forecasting under-represented extreme values of the target variable. These findings are useful for future ML based turbulence models to evaluate their model predictability and provide guidance to systematically generate diversified high-fidelity simulation database.

physics.flu-dyn↗

arXiv2019

Artificial electromagnetic field with cold atoms in two-dimensional optical lattice

We propose that an artificial electromagnetic field can be engineered in the context of cold fermionic atoms that are coupled to a cavity mode via two-photon processes in a two-dimensional optical lattice. There is a standing-wave pump laser inducing the electric effect in one spatial direction, and a second running-wave laser beam generates the magnetic flux perpendicular to the lattice plane. In the static scenario, the bulk spectrum resembles the fractal structure of the Hofstadter butterfly and the edge mode spectrum indicates the occurrence of the quantum Hall phase. The Keldysh formulism is utilized to capture the time evolution. The back action between atoms and cavity field gives a picture of the time-dependent non-equilibrium dynamics. We find that the spontaneous emergence of the artificial electromagnetic field stimulates the Hall current, and the superradiant cavity field emerges without pump threshold. Moreover, the magnetic flux not only modify the bulk topology, but also drives a series of dynamical phase transitions.

quant-ph↗