arXiv Science⌕ Search

arXiv subjects

Ye Tian

Publications and source records attributed to Ye Tian.

At least 199 records · Page 11Linked to original sources
arXiv2023

RestGPT: Connecting Large Language Models with Real-World RESTful APIs

Tool-augmented large language models (LLMs) have achieved remarkable progress in tackling a broad range of tasks. However, existing methods are mainly restricted to specifically designed tools and fail to fulfill complex instructions, having great limitations when confronted with real-world scenarios. In this paper, we explore a more realistic scenario by connecting LLMs with RESTful APIs, which adhere to the widely adopted REST software architectural style for web service development. To address the practical challenges of tackling complex instructions, we propose RestGPT, which exploits the power of LLMs and conducts a coarse-to-fine online planning mechanism to enhance the abilities of task decomposition and API selection. RestGPT also contains an API executor tailored for calling RESTful APIs, which can meticulously formulate parameters and parse API responses. To fully evaluate the performance of RestGPT, we propose RestBench, a high-quality benchmark which consists of two real-world scenarios and human-annotated instructions with gold solution paths. Experiments show that RestGPT is able to achieve impressive results in complex tasks and has strong robustness, which paves a new way towards AGI. RestGPT and RestBench is publicly available at https://restgpt.github.io/.

cs.CL↗
arXiv2023

EQ-Net: Elastic Quantization Neural Networks

Current model quantization methods have shown their promising capability in reducing storage space and computation complexity. However, due to the diversity of quantization forms supported by different hardware, one limitation of existing solutions is that usually require repeated optimization for different scenarios. How to construct a model with flexible quantization forms has been less studied. In this paper, we explore a one-shot network quantization regime, named Elastic Quantization Neural Networks (EQ-Net), which aims to train a robust weight-sharing quantization supernet. First of all, we propose an elastic quantization space (including elastic bit-width, granularity, and symmetry) to adapt to various mainstream quantitative forms. Secondly, we propose the Weight Distribution Regularization Loss (WDR-Loss) and Group Progressive Guidance Loss (GPG-Loss) to bridge the inconsistency of the distribution for weights and output logits in the elastic quantization space gap. Lastly, we incorporate genetic algorithms and the proposed Conditional Quantization-Aware Accuracy Predictor (CQAP) as an estimator to quickly search mixed-precision quantized neural networks in supernet. Extensive experiments demonstrate that our EQ-Net is close to or even better than its static counterparts as well as state-of-the-art robust bit-width methods. Code can be available at \href{https://github.com/xuke225/EQ-Net.git}{https://github.com/xuke225/EQ-Net}.

cs.CV↗
arXiv2023

Improving Social Media Popularity Prediction with Multiple Post Dependencies

Social Media Popularity Prediction has drawn a lot of attention because of its profound impact on many different applications, such as recommendation systems and multimedia advertising. Despite recent efforts to leverage the content of social media posts to improve prediction accuracy, many existing models fail to fully exploit the multiple dependencies between posts, which are important to comprehensively extract content information from posts. To tackle this problem, we propose a novel prediction framework named Dependency-aware Sequence Network (DSN) that exploits both intra- and inter-post dependencies. For intra-post dependency, DSN adopts a multimodal feature extractor with an efficient fine-tuning strategy to obtain task-specific representations from images and textual information of posts. For inter-post dependency, DSN uses a hierarchical information propagation method to learn category representations that could better describe the difference between posts. DSN also exploits recurrent networks with a series of gating layers for more flexible local temporal processing abilities and multi-head attention for long-term dependencies. The experimental results on the Social Media Popularity Dataset demonstrate the superiority of our method compared to existing state-of-the-art models.

cs.MM↗
arXiv2023

Quantum simulation of topological zero modes on a 41-qubit superconducting processor

Quantum simulation of different exotic topological phases of quantum matter on a noisy intermediate-scale quantum (NISQ) processor is attracting growing interest. Here, we develop a one-dimensional 43-qubit superconducting quantum processor, named as Chuang-tzu, to simulate and characterize emergent topological states. By engineering diagonal Aubry-Andr$\acute{\mathrm{e}}$-Harper (AAH) models, we experimentally demonstrate the Hofstadter butterfly energy spectrum. Using Floquet engineering, we verify the existence of the topological zero modes in the commensurate off-diagonal AAH models, which have never been experimentally realized before. Remarkably, the qubit number over 40 in our quantum processor is large enough to capture the substantial topological features of a quantum system from its complex band structure, including Dirac points, the energy gap's closing, the difference between even and odd number of sites, and the distinction between edge and bulk states. Our results establish a versatile hybrid quantum simulation approach to exploring quantum topological systems in the NISQ era.

quant-ph↗
arXiv2023

Designing Novel Cognitive Diagnosis Models via Evolutionary Multi-Objective Neural Architecture Search

Cognitive diagnosis plays a vital role in modern intelligent education platforms to reveal students' proficiency in knowledge concepts for subsequent adaptive tasks. However, due to the requirement of high model interpretability, existing manually designed cognitive diagnosis models hold too simple architectures to meet the demand of current intelligent education systems, where the bias of human design also limits the emergence of effective cognitive diagnosis models. In this paper, we propose to automatically design novel cognitive diagnosis models by evolutionary multi-objective neural architecture search (NAS). Specifically, we observe existing models can be represented by a general model handling three given types of inputs and thus first design an expressive search space for the NAS task in cognitive diagnosis. Then, we propose multi-objective genetic programming (MOGP) to explore the NAS task's search space by maximizing model performance and interpretability. In the MOGP design, each architecture is transformed into a tree architecture and encoded by a tree for easy optimization, and a tailored genetic operation based on four sub-genetic operations is devised to generate offspring effectively. Besides, an initialization strategy is also suggested to accelerate the convergence by evolving half of the population from existing models' variants. Experiments on two real-world datasets demonstrate that the cognitive diagnosis models searched by the proposed approach exhibit significantly better performance than existing models and also hold as good interpretability as human-designed models.

cs.NE↗
arXiv2023

Universal Quantum Optimization with Cold Atoms in an Optical Cavity

Cold atoms in an optical cavity have been widely used for quantum simulations of many-body physics, where the quantum control capability has been advancing rapidly in recent years. Here, we show the atom cavity system is universal for quantum optimization with arbitrary connectivity. We consider a single-mode cavity and develop a Raman coupling scheme by which the engineered quantum Hamiltonian for atoms directly encodes number partition problems (NPPs). The programmability is introduced by placing the atoms at different positions in the cavity with optical tweezers. The NPP solution is encoded in the ground state of atomic qubits coupled through a photonic cavity mode, that can be reached by adiabatic quantum computing (AQC). We construct an explicit mapping for the 3-SAT and vertex cover problems to be efficiently encoded by the cavity system, which costs linear overhead in the number of atomic qubits. The atom cavity encoding is further extended to quadratic unconstrained binary optimization (QUBO) problems. The encoding protocol is optimal in the cost of atom number scaling with the number of binary degrees of freedom of the computation problem. Our theory implies the atom cavity system is a promising quantum optimization platform searching for practical quantum advantage.

quant-ph↗
arXiv2023

A Macro-Micro Approach to Reconstructing Vehicle Trajectories on Multi-Lane Freeways with Lane Changing

Vehicle trajectories can offer the most precise and detailed depiction of traffic flow and serve as a critical component in traffic management and control applications. Various technologies have been applied to reconstruct vehicle trajectories from sparse fixed and mobile detection data. However, existing methods predominantly concentrate on single-lane scenarios and neglect lane-changing (LC) behaviors that occur across multiple lanes, which limit their applicability in practical traffic systems. To address this research gap, we propose a macro-micro approach for reconstructing complete vehicle trajectories on multi-lane freeways, wherein the macro traffic state information and micro driving models are integrated to overcome the restrictions imposed by lane boundary. Particularly, the macroscopic velocity contour maps are established for each lane to regulate the movement of vehicle platoons, meanwhile the velocity difference between adjacent lanes provide valuable criteria for guiding LC behaviors. Simultaneously, the car-following models are extended from micro perspective to supply lane-based candidate trajectories and define the plausible range for LC positions. Later, a two-stage trajectory fusion algorithm is proposed to jointly infer both the car-following and LC behaviors, in which the optimal LC positions is identified and candidate trajectories are adjusted according to their weights. The proposed framework was evaluated using NGSIM dataset, and the results indicated a remarkable enhancement in both the accuracy and smoothness of reconstructed trajectories, with performance indicators reduced by over 30% compared to two representative reconstruction methods. Furthermore, the reconstruction process effectively reproduced LC behaviors across contiguous lanes, adding to the framework's comprehensiveness and realism.

math.OC↗
arXiv2023

Tunable Coupling Architectures with Capacitively Connecting Pads for Large-Scale Superconducting Multi-Qubit Processors

We have proposed and experimentally verified a tunable inter-qubit coupling scheme for large-scale integration of superconducting qubits. The key feature of the scheme is the insertion of connecting pads between qubit and tunable coupling element. In such a way, the distance between two qubits can be increased considerably to a few millimeters, leaving enough space for arranging control lines, readout resonators and other necessary structures. The increased inter-qubit distance provides more wiring space for flip-chip process and reduces crosstalk between qubits and from control lines to qubits. We use the term Tunable Coupler with Capacitively Connecting Pad (TCCP) to name the tunable coupling part that consists of a transmon coupler and capacitively connecting pads. With the different placement of connecting pads, different TCCP architectures can be realized. We have designed and fabricated a few multi-qubit devices in which TCCP is used for coupling. The measured results show that the performance of the qubits coupled by the TCCP, such as $T_1$ and $T_2$, was similar to that of the traditional transmon qubits without TCCP. Meanwhile, our TCCP also exhibited a wide tunable range of the effective coupling strength and a low residual ZZ interaction between the qubits by properly tuning the parameters on the design. Finally, we successfully implemented an adiabatic CZ gate with TCCP. Furthermore, by introducing TCCP, we also discuss the realization of the flip-chip process and tunable coupling qubits between different chips.

quant-ph↗
arXiv2023

Quantum simulation of Hawking radiation and curved spacetime with a superconducting on-chip black hole

Hawking radiation is one of the quantum features of a black hole that can be understood as a quantum tunneling across the event horizon of the black hole, but it is quite difficult to directly observe the Hawking radiation of an astrophysical black hole. Here, we report a fermionic lattice-model-type realization of an analogue black hole by using a chain of 10 superconducting transmon qubits with interactions mediated by 9 transmon-type tunable couplers. The quantum walks of quasi-particle in the curved spacetime reflect the gravitational effect near the black hole, resulting in the behaviour of stimulated Hawking radiation, which is verified by the state tomography measurement of all 7 qubits outside the horizon. In addition, the dynamics of entanglement in the curved spacetime is directly measured. Our results would stimulate more interests to explore the related features of black holes using the programmable superconducting processor with tunable couplers.

quant-ph↗
arXiv2023

Coordinated Frequency-Constrained Stochastic Economic Dispatch for Integrated Transmission and Distribution System via Distributed Optimization

When large-scale uncertain centralized and distributed renewable energy sources are connected to a power system, separate dispatching of the transmission power system (TPS) and the active distribution network (ADN) will lower the network security and frequency security of the system. To address these problems, this paper proposes a coordinated frequency-constrained stochastic economic dispatch (CFC-SED) model for an integrated transmission and distribution (ITD) system. In this model, the dynamic frequency security constraints and network security constraints of the ITD system are constructed, and the joint chance constraints are adopted to handle the uncertainty. Then, the control parameters of inverter-based resources, the base point power, and the regulation reserve of all dispatchable resources in the ITD system are jointly optimized for the minimum operating cost. TPS and ADNs can deliver base point power bidirectionally and provide frequency regulation support bidirectionally, which extend the existing reserve assumption in ITD dispatch and enhance the operational security of the ITD system. Moreover, based on the alternating direction of multiplier algorithm, a two-layer distributed optimization framework is proposed to solve the CFC-SED model. Case studies show that the CFC-SED model can fully utilize the potential of multiple regulation resources to improve the security performance of the ITD system, and TPS and ADNs can be coordinated efficiently through the proposed distributed optimization framework.

eess.SY↗
arXiv2023

Machine learning aided atomic structure identification of interfacial ionic hydrates from atomic force microscopy images

Relevant to broad applied fields and natural processes, interfacial ionic hydrates has been widely studied by ultrahigh-resolution atomic force microscopy (AFM). However, the complex relationship between AFM signal and the investigated system makes it difficult to determine the atomic structure of such complex system from AFM images alone. Using machine learning, we achieved precise identification of the atomic structures of interfacial water/ionic hydrates based on AFM images, including the position of each atom and the orientation of water molecules. Furthermore, it was found that structure prediction of ionic hydrates can be achieved cost-effectively by transfer learning using neural network (NN) trained with easily available interfacial water data. Thus, this work provides an efficient and economical methodology which not only opens up avenues to determine atomic structures of more complex systems from AFM images, but may also help to interpret other science-wide studies involving sophisticated experimental results.

cond-mat.mes-hall↗
arXiv2023

Joint Chance-Constrained Economic Dispatch Involving Joint Optimization of Frequency-related Inverter Control and Regulation Reserve Allocation

The issues of uncertainty and frequency security could become significantly serious in power systems with the high penetration of volatile inverter-based renewables (IBRs). These issues make it necessary to consider the uncertainty and frequency-related constraints in the economic dispatch (ED) programs. However, existing ED studies rarely proactively optimize the control parameters of inverter-based resources related to fast regulation (e.g., virtual inertia and droop coefficients) in cooperation with other dispatchable resources to improve the system frequency security and dispatch reliability. This paper first proposes a joint chance-constrained economic dispatch model that jointly optimizes the frequency-related inverter control, the system up/down reserves, and base-point power for the minimal total operational cost. In the proposed model, multiple dispatchable resources including thermal units, dispatchable IBRs and energy storage are considered, and the (virtual) inertias, the regulation reserve allocations and the base-point power are coordinated. To ensure the system reliability, the joint chance-constraint formulation is also adopted. Additionally, since the traditional sample average approximation (SAA) method cost much computational burden, a novel mix-SAA (MSAA) method is proposed to transform the original intractable model into a linear model that can be efficiently solved via commercial solvers. The case studies validated the satisfactory efficacy of the proposed ED model and the efficiency of the MSAA.

eess.SY↗
arXiv2023

How social influence affects the wisdom of crowds in influence networks

A long-standing debate is whether social influence improves the collective wisdom of a crowd or undermines it. This paper addresses this question based on a naive learning setting in influence systems theory: in our models individuals evolve their estimates of an unknown truth according to the weighted-average opinion dynamics. A formal mathematization is provided with rigorous theoretical analysis. We obtain various conditions for improving, optimizing and undermining the crowd accuracy, respectively. We prove that if the wisdom of finite-size group is improved, then the collective estimate converges to the truth as group size increases, provided individuals' variances are finite. We show that whether social influence improves or undermines the wisdom is determined by the social power allocation of the influence system: if the influence system allocates relatively larger social power to relatively more accurate individuals, it improves the wisdom; on the contrary, if the influence system assigns less social power to more accurate individuals, it undermines the wisdom. At a population level, individuals' susceptibilities to interpersonal influence and network centralities are both crucial. To improve the wisdom, more accurate individuals should be less susceptible and have larger network centralities. Particularly, in democratic influence networks, if relatively more accurate individuals are relatively less susceptible, the wisdom is improved; if more accurate individuals are more susceptible, the wisdom is undermined, which is consistent with the reported empirical evidence. Our investigation provides a theoretical framework for understanding the role social influence plays in the emergence of collective wisdom.

cs.SI↗
arXiv2023

The Solenoidal Large Intensity Device (SoLID) for JLab 12 GeV

The Solenoidal Large Intensity Device (SoLID) is a new experimental apparatus planned for Hall A at the Thomas Jefferson National Accelerator Facility (JLab). SoLID will combine large angular and momentum acceptance with the capability to handle very high data rates at high luminosity. With a slate of approved high-impact physics experiments, SoLID will push JLab to a new limit at the QCD intensity frontier that will exploit the full potential of its 12 GeV electron beam. In this paper, we present an overview of the rich physics program that can be realized with SoLID, which encompasses the tomography of the nucleon in 3-D momentum space from Semi-Inclusive Deep Inelastic Scattering (SIDIS), expanding the phase space in the search for new physics and novel hadronic effects in parity-violating DIS (PVDIS), a precision measurement of $J/ψ$ production at threshold that probes the gluon field and its contribution to the proton mass, tomography of the nucleon in combined coordinate and momentum space with deep exclusive reactions, and more. To meet the challenging requirements, the design of SoLID described here takes full advantage of recent progress in detector, data acquisition and computing technologies. In addition, we outline potential experiments beyond the currently approved program and discuss the physics that could be explored should upgrades of CEBAF become a reality in the future.

nucl-ex↗
arXiv2023

Improving Across-Dataset Brain Tissue Segmentation Using Transformer

Brain tissue segmentation has demonstrated great utility in quantifying MRI data through Voxel-Based Morphometry and highlighting subtle structural changes associated with various conditions within the brain. However, manual segmentation is highly labor-intensive, and automated approaches have struggled due to properties inherent to MRI acquisition, leaving a great need for an effective segmentation tool. Despite the recent success of deep convolutional neural networks (CNNs) for brain tissue segmentation, many such solutions do not generalize well to new datasets, which is critical for a reliable solution. Transformers have demonstrated success in natural image segmentation and have recently been applied to 3D medical image segmentation tasks due to their ability to capture long-distance relationships in the input where the local receptive fields of CNNs struggle. This study introduces a novel CNN-Transformer hybrid architecture designed for brain tissue segmentation. We validate our model's performance across four multi-site T1w MRI datasets, covering different vendors, field strengths, scan parameters, time points, and neuropsychiatric conditions. In all situations, our model achieved the greatest generality and reliability. Out method is inherently robust and can serve as a valuable tool for brain-related T1w MRI studies. The code for the TABS network is available at: https://github.com/raovish6/TABS.

eess.IV↗
arXiv2022

DUNE: Improving Accuracy for Sketch-INT Network Measurement Systems

In-band Network Telemetry (INT) and sketching algorithms are two promising directions for measuring network traffics in real time. To combine sketch with INT and preserve their advantages, a representative approach is to use INT to send a switch sketch in small pieces (called sketchlets) to end-host for reconstructing an identical sketch. However, in this paper, we reveal that when naively selecting buckets to sketchlets, the end-host reconstructed sketch is inaccurate. To overcome this problem, we present DUNE, an innovative sketch-INT network measurement system. DUNE incorporates two key innovations: First, we design a novel scatter sketchlet that is more efficient in transferring measurement data by allowing a switch to select individual buckets to add to sketchlets; Second, we propose lightweight data structures for tracing "freshness" of the sketch buckets, and present algorithms for smartly selecting buckets that contain valuable measurement data to send to end-host. We theoretically prove the effectiveness of our proposed methods, and implement a prototype on commodity programmable switch. The results of extensive experiments driven by real-world traffics on DUNE suggest that our proposed system can substantially improve the measurement accuracy at a trivial cost.

cs.NI↗
arXiv2022

Reciprocal phase transition-enabled electro-optic modulation

Electro-optic (EO) modulation is a well-known and essential topic in the field of communications and sensing. Its ultrahigh efficiency is unprecedentedly desired in the current green and data era. However, dramatically increasing the modulation efficiency is difficult due to the monotonic mapping relationship between the electrical signal and modulated optical signal. Here, a new mechanism termed phase-transition EO modulation is revealed from the reciprocal transition between two distinct phase planes arising from the bifurcation. Remarkably, a monolithically integrated mode-locked laser (MLL) is implemented as a prototype. A 24.8-GHz radio-frequency signal is generated and modulated, achieving a modulation energy efficiency of 3.06 fJ/bit improved by about four orders of magnitude and a contrast ratio exceeding 50 dB. Thus, MLL-based phase-transition EO modulation is characterised by ultrahigh modulation efficiency and ultrahigh contrast ratio, as experimentally proved in radio-over-fibre and underwater acoustic-sensing systems. This phase-transition EO modulation opens a new avenue for green communication and ubiquitous connections.

physics.optics↗
arXiv2022

Deep Maxout Network Gaussian Process

Study of neural networks with infinite width is important for better understanding of the neural network in practical application. In this work, we derive the equivalence of the deep, infinite-width maxout network and the Gaussian process (GP) and characterize the maxout kernel with a compositional structure. Moreover, we build up the connection between our deep maxout network kernel and deep neural network kernels. We also give an efficient numerical implementation of our kernel which can be adapted to any maxout rank. Numerical results show that doing Bayesian inference based on the deep maxout network kernel can lead to competitive results compared with their finite-width counterparts and deep neural network kernels. This enlightens us that the maxout activation may also be incorporated into other infinite-width neural network structures such as the convolutional neural network (CNN).

stat.ML↗