Searcharxiv⌕ Search

arXiv subjects

Siyuan Lu

Publications and source records attributed to Siyuan Lu.

At least 37 records · Page 2Linked to original sources
arXiv2024

PreAct: Prediction Enhances Agent's Planning Ability

Addressing the disparity between forecasts and actual results can enable individuals to expand their thought processes and stimulate self-reflection, thus promoting accurate planning. In this research, we present **PreAct**, an agent framework that integrates **pre**diction, **rea**soning, and **act**ion. By utilizing the information derived from predictions, the large language model (LLM) agent can provide a wider range and more strategically focused reasoning. This leads to more efficient actions that aid the agent in accomplishing intricate tasks. Our experimental results show that PreAct surpasses the ReAct method in completing complex tasks and that PreAct's performance can be further improved when paired with other memory or selection strategy techniques. We presented the model with varying quantities of historical predictions and discovered that these predictions consistently enhance LLM planning.The variances in single-step reasoning between PreAct and ReAct indicate that PreAct indeed has benefits in terms of diversity and strategic orientation over ReAct.

cs.CL↗
arXiv2024

Bell state generation and CNOT operation using on-demand identical photons from shape-controlled spatially ordered quantum dots

Fault tolerant on-chip photonic quantum computation is enormously helped by (a) deterministic generation of the needed thousands to millions of photon qubits from (b) quantum emitters in designed spatially ordered arrays to enable networks for implementing many-qubit logic circuits. Scaling up photonic quantum information processing systems has, however, been prevented by the lack of such quantum emitters until the demonstration of the platform of mesa-top single quantum dots (MTSQDs) -- controlled shape, size, and volume single QD -- located in designed regular arrays. Here we demonstrate 2 qubit CNOT gate operation -- a universal gate necessary to enable quantum circuits of arbitrary complexity -- in polarization basis using photons emitted from individual MTSQDs. A Bell state fidelity of 0.825$\pm$0.010 is achieved with two photon interference (TPI) visibility of 0.947$\pm$0.0015 at 4K without Purcell enhancement. The results make a strong case for developing MTSQD arrays for utility scale optical quantum information processing platforms.

quant-ph↗
arXiv2024

Interior $C^2$ estimate for Hessian quotient equation in general dimension

In this paper, we study the interior $C^2$ regularity problem for the Hessian quotient equation $\left(\frac{σ_n}{σ_k}\right)(D^2u)=f$. We give a complete answer to this longstanding problem: for $k=n-1,n-2$, we establish an interior $C^2$ estimate; for $k\leq n-3$, we show that interior $C^2$ estimate fails by finding a singular solution.

math.AP↗
arXiv2023

Large-Area Spatially Ordered Mesa Top Single Quantum Dots: Suitable Single Photon Emitters for On-Chip Integrated Quantum Information Processing Platforms

Realization of the long sought on-chip scalable photonic quantum information processing networks has been thwarted by the absence of spatially-ordered and scalable on-demand single photon emitters with emission figures-of-merit exceeding the required thresholds across large numbers. The positioning must meet the required degree of accuracy that enables fabricating their interconnection to create the desired functional network. Here we report on the realization of large-area spatially-ordered arrays of mesa-top single quantum dots (MTSQDs) that are demonstrated [1] to be on-demand single photon emitters with characteristics that meet the requirements for implementing quantum photonic circuits/platforms aimed at quantum key distribution, linear optical quantum computing, simulations of quantum many-body problems, and metrology/sensing. The reported GaAs/InGaAs/GaAs MTSQD arrays, grown via SESRE (substrate-encoded size-reducing epitaxy) are in multiple arrays of up to 100x100 with 5um pitch, across a centimeter radius area. We show illustrative large-area images of the emission intensity (brightness) and color-coded wavelength distribution exhibiting ~3.35nm standard deviation. Scanning transmission electron microscopy shows a remarkable control on the QD location to within ~3nm accuracy laterally and ~1nm vertically. The primary remaining challenge is the control on the uniformity of the currently wet-chemically etched as-patterned nanomesa lateral size across the substrate, a surmountable technical issue. Thus, SESRE offers the most promising approach to realizing on-chip scalable spatially-ordered arrays of on-demand bright single quantum emitters meeting the figures-of-merit required for on-chip fully integrated quantum photonic circuit platforms-monolithic (such as based upon AlGaAs on insulator) or hybrid that leverage the silicon-on-insulator (SOI) photonic integrated circuit (PIC).

quant-ph↗
arXiv2023

On the asymptotic Plateau problem in hyperbolic space

In this paper, we solve the asymptotic Plateau problem in hyperbolic space for constant $σ_{n-1}$ curvature, i.e. the existence of a complete hypersurface in $\mathbb{H}^{n+1}$ satisfying $σ_{n-1}(κ)=σ\in (0,n)$ with a prescribed asymptotic boundary $Γ$. The key ingredient is the curvature estimates. Previously, this is only known for $σ_0<σ<n$, where $σ_0$ is a positive constant.

math.DG↗
arXiv2022

Fast and Accurate FSA System Using ELBERT: An Efficient and Lightweight BERT

With the development of deep learning and Transformer-based pre-trained models like BERT, the accuracy of many NLP tasks has been dramatically improved. However, the large number of parameters and computations also pose challenges for their deployment. For instance, using BERT can improve the predictions in the financial sentiment analysis (FSA) task but slow it down, where speed and accuracy are equally important in terms of profits. To address these issues, we first propose an efficient and lightweight BERT (ELBERT) along with a novel confidence-window-based (CWB) early exit mechanism. Based on ELBERT, an innovative method to accelerate text processing on the GPU platform is developed, solving the difficult problem of making the early exit mechanism work more effectively with a large input batch size. Afterward, a fast and high-accuracy FSA system is built. Experimental results show that the proposed CWB early exit mechanism achieves significantly higher accuracy than existing early exit methods on BERT under the same computation cost. By using this acceleration method, our FSA system can boost the processing speed by nearly 40 times to over 1000 texts per second with sufficient accuracy, which is nearly twice as fast as FastBERT, thus providing a more powerful text processing capability for modern trading systems.

cs.CL↗
arXiv2022

Rethinking WMMSE: Can Its Complexity Scale Linearly With the Number of BS Antennas?

Precoding design for maximizing weighted sum-rate (WSR) is a fundamental problem for downlink of massive multi-user multiple-input multiple-output (MU-MIMO) systems. It is well-known that this problem is generally NP-hard due to the presence of multi-user interference. The weighted minimum mean-square error (WMMSE) algorithm is a popular approach for WSR maximization. However, its computational complexity is cubic in the number of base station (BS) antennas, which is unaffordable when the BS is equipped with a large antenna array. In this paper, we consider the WSR maximization problem with either a sum-power constraint (SPC) or per-antenna power constraints (PAPCs). For the former, we prove that any nontrivial stationary point must have a low-dimensional subspace structure, and then propose a reduced-WMMSE (R-WMMSE) with linear complexity by exploiting the solution structure. For the latter, we propose a linear-complexity WMMSE approach, named PAPC-WMMSE, by using a novel recursive design of the algorithm. Both R-WMMSE and PAPC-WMMSE have simple closed-form updates and guaranteed convergence to stationary points. Simulation results verify the efficacy of the proposed designs, especially the much lower complexity as compared to the state-of-the-art approaches for massive MU-MIMO systems.

eess.SP↗
arXiv2021

On chip scalable highly pure and indistinguishable single photon sources in ordered arrays: Path to Quantum Optical Circuits

Realization of quantum optical circuits is at the heart of quantum photonic information processing. A long-standing obstacle however has been the absence of a platform of single photon sources (SPSs) that simultaneously satisfies the following required characteristics: spatially ordered SPS arrays that produce, on-demand, highly pure, and indistinguishable single photons with sufficiently uniform emission characteristics across the array, needed for controlled interference between photons from distinct sources to enable functional quantum optical networks. Here we report on such a platform of SPSs based upon a novel class of epitaxial quantum dots. Under resonant excitation, the SPSs (without Purcell enhancement) show single photon purity >99% (g(2)(0) ~ 0.015), high two-photon Hong-Ou-Mandel interference visibilities of 0.82 $\pm$ 0.03 (at 11.5K), and spectral nonuniformity <3nm - within established locally tunable technology. Our platform of SPSs paves the path to creating on-chip scalable quantum photonic systems.

quant-ph↗
arXiv2021

On the $σ_2$-Nirenberg problem on $\mathbb{S}^2$

We establish theorems on the existence and compactness of solutions to the $σ_2$-Nirenberg problem on the standard sphere $\mathbb S^2$. A first significant ingredient, a Liouville type theorem for the associated fully nonlinear Möbius invariant elliptic equations, was established in an earlier paper of ours. Our proof of the existence and compactness results requires a number of additional crucial ingredients which we prove in this paper: A Liouville type theorem for the associated fully nonlinear Möbius invariant degenerate elliptic equations, a priori estimates of first and second order derivatives of solutions to the $σ_2$-Nirenberg problem, and a Bôcher type theorem for the associated fully nonlinear Möbius invariant elliptic equations. Given these results, we are able to complete a fine analysis of a sequence of blow-up solutions to the $σ_2$-Nirenberg problem. In particular, we prove that there can be at most one blow-up point for such a blow-up sequence of solutions. This, together with a Kazdan-Warner type identity, allows us to prove $L^\infty$ a priori estimates for solutions of the $σ_2$-Nirenberg problem under some simple generic hypothesis. The higher derivative estimates then follow from classical estimates of Nirenberg and Schauder. In turn, the existence of solutions to the $σ_2$-Nirenberg problem is obtained by an application of the by now standard degree theory for second order fully nonlinear elliptic operators.

math.AP↗
arXiv2021

Elbert: Fast Albert with Confidence-Window Based Early Exit

Despite the great success in Natural Language Processing (NLP) area, large pre-trained language models like BERT are not well-suited for resource-constrained or real-time applications owing to the large number of parameters and slow inference speed. Recently, compressing and accelerating BERT have become important topics. By incorporating a parameter-sharing strategy, ALBERT greatly reduces the number of parameters while achieving competitive performance. Nevertheless, ALBERT still suffers from a long inference time. In this work, we propose the ELBERT, which significantly improves the average inference speed compared to ALBERT due to the proposed confidence-window based early exit mechanism, without introducing additional parameters or extra training overhead. Experimental results show that ELBERT achieves an adaptive inference speedup varying from 2$\times$ to 10$\times$ with negligible accuracy degradation compared to ALBERT on various datasets. Besides, ELBERT achieves higher accuracy than existing early exit methods used for accelerating BERT under the same computation cost. Furthermore, to understand the principle of the early exit mechanism, we also visualize the decision-making process of it in ELBERT.

cs.CL↗
arXiv2020

A Liouville theorem for Möbius invariant equations

In this paper we classify Möbius invariant differential operators of second order in two dimensional Euclidean space, and establish a Liouville type theorem for general Möbius invariant elliptic equations.

math.AP↗
arXiv2020

PAIRS AutoGeo: an Automated Machine Learning Framework for Massive Geospatial Data

An automated machine learning framework for geospatial data named PAIRS AutoGeo is introduced on IBM PAIRS Geoscope big data and analytics platform. The framework simplifies the development of industrial machine learning solutions leveraging geospatial data to the extent that the user inputs are minimized to merely a text file containing labeled GPS coordinates. PAIRS AutoGeo automatically gathers required data at the location coordinates, assembles the training data, performs quality check, and trains multiple machine learning models for subsequent deployment. The framework is validated using a realistic industrial use case of tree species classification. Open-source tree species data are used as the input to train a random forest classifier and a modified ResNet model for 10-way tree species classification based on aerial imagery, which leads to an accuracy of $59.8\%$ and $81.4\%$, respectively. This use case exemplifies how PAIRS AutoGeo enables users to leverage machine learning without extensive geospatial expertise.

cs.CV↗
arXiv2020

A clustered Gaussian process model for computer experiments

A Gaussian process has been one of the important approaches for emulating computer simulations. However, the stationarity assumption for a Gaussian process and the intractability for large-scale dataset limit its availability in practice. In this article, we propose a clustered Gaussian process model which segments the input data into multiple clusters, in each of which a Gaussian process model is performed. The stochastic expectation-maximization is employed to efficiently fit the model. In our simulations as well as a real application to solar irradiance emulation, our proposed method had smaller mean square errors than its main competitors, with competitive computation time, and provides valuable insights from data by discovering the clusters. An R package for the proposed methodology is provided in an open repository.

stat.ME↗
arXiv2020

Rigidity of Riemannian Penrose inequality with corners and its implications

Motivated by the rigidity case in the localized Riemannian Penrose inequality, we show that suitable singular metrics attaining the optimal value in the Riemannian Penrose inequality is necessarily smooth in properly specified coordinates. If applied to hypersurfaces enclosing the horizon in a spatial Schwarzschild manifold, the result gives the rigidity of isometric hypersurfaces with the same mean curvature.

math.DG↗