arXiv subjects

Jun Gao

Publications and source records attributed to Jun Gao.

At least 181 records · Page 10Linked to original sources

arXiv2023

Benchmarking Large Language Models with Augmented Instructions for Fine-grained Information Extraction

Information Extraction (IE) is an essential task in Natural Language Processing. Traditional methods have relied on coarse-grained extraction with simple instructions. However, with the emergence of Large Language Models (LLMs), there is a need to adapt IE techniques to leverage the capabilities of these models. This paper introduces a fine-grained IE benchmark dataset tailored for LLMs, employing augmented instructions for each information type, which includes task descriptions, extraction rules, output formats, and examples. Through extensive evaluations, we observe that encoder-decoder models, particularly T5 and FLAN-T5, perform well in generalizing to unseen information types, while ChatGPT exhibits greater adaptability to new task forms. Our results also indicate that performance is not solely dictated by model scale, and highlight the significance of architecture, data diversity, and learning techniques. This work paves the way for a more refined and versatile utilization of LLMs in Information Extraction.

cs.CL↗

arXiv2023

Bayesian Inference of Supernova Neutrino Spectra with Multiple Detectors

We implement the Bayesian inference to retrieve energy spectra of all neutrinos from a galactic core-collapse supernova (CCSN). To achieve high statistics and full sensitivity to all flavours of neutrinos, we adopt a combination of several reaction channels from different large-scale neutrino observatories, namely inverse beta decay on proton and elastic scattering on electron from Hyper-Kamiokande (Hyper-K), charged current absorption on Argon from Deep Underground Neutrino Experiment (DUNE) and coherent elastic scattering on Lead from RES-NOVA. Assuming no neutrino oscillation or specific oscillation models, we obtain mock data for each channel through Poisson processes with the predictions, for a typical source distance of 10 kpc in our Galaxy, and then evaluate the probability distributions for all spectral parameters of theoretical neutrino spectrum model with Bayes' theorem. Although the results for either the electron-neutrinos or electron-antineutrinos reserve relatively large uncertainties (according to the neutrino mass hierarchy), a precision of a few percent (i.e., $\pm 1 \% \sim \pm 4 \%$ at a credible interval of $2 σ$) is achieved for primary spectral parameters (e.g., mean energy and total emitted energy) of other neutrino species. Moreover, the correlation coefficients between different parameters are computed as well and interesting patterns are found. Especially, the mixing-induced correlations are sensitive to the neutrino mass hierarchy, which potentially makes it a brand new probe to determine the neutrino mass hierarchy in the detection of galactic supernova neutrinos. Finally, we discuss the origin of such correlation patterns and perspectives for further improvement on our results.

hep-ph↗

arXiv2023

Bayesian-Boosted MetaLoc: Efficient Training and Guaranteed Generalization for Indoor Localization

Existing localization approaches utilizing environment-specific channel state information (CSI) excel under specific environment but struggle to generalize across varied environments. This challenge becomes even more pronounced when confronted with limited training data. To address these issues, we present the Bayes-Optimal Meta-Learning for Localization (BOML-Loc) framework, inspired by the PAC-Optimal Hyper-Posterior (PACOH) algorithm. Improving on our earlier MetaLoc~\cite{MetaLoc}, BOML-Loc employs a Bayesian approach, reducing the need for extensive training, lowering overfitting risk, and offering per-test-point uncertainty estimation. Even with very limited training tasks, BOML-Loc guarantees robust localization and impressive generalization. In both LOS and NLOS environments with site-surveyed data, BOML-Loc surpasses existing models, demonstrating enhanced localization accuracy, generalization abilities, and reduced overfitting in new and previously unseen environments.

eess.SP↗

arXiv2023

MetaLoc: Learning to Learn Wireless Localization

Existing localization methods that intensively leverage the environment-specific received signal strength (RSS) or channel state information (CSI) of wireless signals are rather accurate in certain environments. However, these methods, whether based on pure statistical signal processing or data-driven approaches, often struggle to generalize to new environments, which results in considerable time and effort being wasted. To address this challenge, we propose MetaLoc, which is the first fingerprinting-based localization framework that leverages the Model-Agnostic Meta-Learning (MAML). Specifically, built on a deep neural network with strong representation capabilities, MetaLoc is trained on historical data sourced from well-calibrated environments, employing a two-loop optimization mechanism to obtain the meta-parameters. These meta-parameters act as the initialization for quick adaptation in new environments, reducing the need for much human effort. The framework introduces two paradigms for the optimization of meta-parameters: a centralized paradigm that simplifies the process by sharing data from all historical environments, and a distributed paradigm that maintains data privacy by training meta-parameters for each specific environment separately. Furthermore, the advanced distributed paradigm modifies the vanilla MAML loss function to ensure that the reduction of loss occurs in a consistent direction across various training domains, thus facilitating faster convergence during training. Our experiments on both synthetic and real datasets demonstrate that MetaLoc outperforms baseline methods in terms of localization accuracy, robustness, and cost-effectiveness. The code and datasets used in this study are publicly available.

eess.SP↗

arXiv2023

Extremal number of graphs from geometric shapes

We study the Turán problem for highly symmetric bipartite graphs arising from geometric shapes and periodic tilings commonly found in nature. 1. The prism $C_{2\ell}^{\square}:=C_{2\ell}\square K_{2}$ is the graph consisting of two vertex disjoint $2\ell$-cycles and a matching pairing the corresponding vertices of these two cycles. We show that for every $\ell\ge 4$, ex$(n,C_{2\ell}^{\square})=Θ(n^{3/2})$. This resolves a conjecture of He, Li and Feng. 2. The hexagonal tiling in honeycomb is one of the most natural structures in the real world. We show that the extremal number of honeycomb graphs has the same order of magnitude as their basic building unit 6-cycles. 3. We also consider bipartite graphs from quadrangulations of the cylinder and the torus. We prove near optimal bounds for both configurations. In particular, our method gives a very short proof of a tight upper bound for the extremal number of the 2-dimensional grid, improving a recent result of Bradač, Janzer, Sudakov and Tomon. Our proofs mix several ideas, including shifting embedding schemes, weighted homomorphism and subgraph counts and asymmetric dependent random choice.

math.CO↗

arXiv2023

Dynamic strain modulation of a nanowire quantum dot compatible with a thin-film lithium niobate photonic platform

The integration of on-demand single photon sources in photonic circuits is a major prerequisite for on-chip quantum applications. Among the various high-quality sources, nanowire quantum dots can be efficiently coupled to optical waveguides because of their preferred emission direction along their growth direction. However, local tuning of the emission properties remains challenging. In this work, we transfer a nanowire quantum dot on a bulk lithium niobate substrate and show that its emission can be dynamically tuned by acousto-optical coupling with surface acoustic waves. The purity of the single photon source is preserved during the strain modulation. We further demonstrate that the transduction is maintained even with a SiO2 encapsulation layer deposited on top of the nanowire acting as the cladding of a photonic circuit. Based on these experimental findings and numerical simulations, we introduce a device architecture consisting of a nanowire quantum dot efficiently coupled to a thin film lithium niobate rib waveguide and strain-tunable by surface acoustic waves.

cond-mat.mes-hall↗

arXiv2023

Flexible Isosurface Extraction for Gradient-Based Mesh Optimization

This work considers gradient-based mesh optimization, where we iteratively optimize for a 3D surface mesh by representing it as the isosurface of a scalar field, an increasingly common paradigm in applications including photogrammetry, generative modeling, and inverse physics. Existing implementations adapt classic isosurface extraction algorithms like Marching Cubes or Dual Contouring; these techniques were designed to extract meshes from fixed, known fields, and in the optimization setting they lack the degrees of freedom to represent high-quality feature-preserving meshes, or suffer from numerical instabilities. We introduce FlexiCubes, an isosurface representation specifically designed for optimizing an unknown mesh with respect to geometric, visual, or even physical objectives. Our main insight is to introduce additional carefully-chosen parameters into the representation, which allow local flexible adjustments to the extracted mesh geometry and connectivity. These parameters are updated along with the underlying scalar field via automatic differentiation when optimizing for a downstream task. We base our extraction scheme on Dual Marching Cubes for improved topological properties, and present extensions to optionally generate tetrahedral and hierarchically-adaptive meshes. Extensive experiments validate FlexiCubes on both synthetic benchmarks and real-world applications, showing that it offers significant improvements in mesh quality and geometric fidelity.

cs.GR↗

arXiv2023

The Erdős distinct subset sums problem in a modular setting

We prove the following variant of the Erdős distinct subset sums problem. Given $t \ge 0$ and sufficiently large $n$, every $n$-element set $A$ whose subset sums are distinct modulo $N=2^n+t$ satisfies $$\max A \ge \Big(\frac{1}{3}-o(1)\Big)N.$$ Furthermore, we provide examples showing that the constant $\frac 13$ is best possible. For small values of $t$, we characterise the structure of all sumset-distinct sets modulo $N=2^n+t$ of cardinality $n$.

math.CO↗

arXiv2023

Constraint for a light charged Higgs boson and its neutral partners from top quark pairs at the LHC

The charged Higgs boson plays an essential role in distinguishing between a wide variety of standard model extensions with multiple Higgs doublets, and has been searched for in various collider experiments. This paper expands our previous work to a broader Higgs mass space with discussions on subsequent issues. We study the prospect of a light charged Higgs boson, produced by top quark pairs at the Large Hadron Collider (LHC), and decaying into a $W$ boson (can be off shell) and a pair of bottom quarks via on-shell production of an intermediate neutral Higgs boson. We reinterpret the cross sections of $WWbb\bar{b}\bar{b}$ final states measured by the ATLAS collaboration at LHC 13 TeV in the presence of the decay chain: $t \rightarrow H^+ b, H^+ \rightarrow W^+ H_i, H_i \rightarrow b \bar{b}$, and H.c., where $H_i$ is a neutral Higgs boson variably lighter than the charged Higgs boson. We find improved agreements with the data and obtain limits on the total branching ratio of the aforementioned decay chain. The limits impose the strongest constraints on the parameter space of type-I two-Higgs-doublet model for most Higgs masses sampled when $H_i$ is the $CP$-odd Higgs boson $A$. We also calculate potential constraints with pseudodata in high-luminosity runs of the LHC.

hep-ph↗

arXiv2023

Nanowire-based Integrated Photonics for Quantum Information and Quantum Sensing

At the core of quantum photonic information processing and sensing, two major building pillarsare single-photon emitters and single-photon detectors. In this review, we systematically summarize the working theory, material platform, fabrication process, and game-changing applications enabled by state-of-the-art quantum dots in nanowire emitters and superconducting nanowire single-photon detectors. Such nanowire-based quantum hardware offers promising properties for modern quantum optics experiments.We highlight several burgeoning quantum photonics applications using nanowires and discuss development trends of integrated quantum photonics. Also, we propose quantum information processing and sensing experiments for the quantum optics community, and future interdisciplinary applications.

quant-ph↗

arXiv2023

NNLL Resummation for Projected Three-Point Energy Correlator

The projected energy correlator measures the energy deposited in multiple detectors as a function of the largest angular distance $x_L = (1 - \cosχ_L)/2$ between detectors. The collinear limit $x_L\to 0$ of the projected energy correlator is particularly interesting for understanding the jet-substructures, while the large logarithms of $x_L$ could potentially spoil the perturbation theory and must be resummed. As a necessary ingredient for its resummation at next-to-next-to-leading logarithmic (NNLL) accuracy, we calculate the two-loop jet functions for the projected three-point energy correlator (E3C), using direct integration method and the parameter space Integration-by-Part (IBP) method. We then present the NNLL resummation for $e^+e^-$ annihilation and an approximate NNLL resummation for $pp\rightarrow jj$ process, where the two-loop hard constant is estimated in the latter case. The convergence is improved and the hadronization effect in the collinear limit is suppressed when considering the ratio of E3C distribution to two-point energy-energy correlator (EEC). Our results show potential in precision determination of strong coupling constant using energy correlators from both $e^+e^-$ data and $pp$ data.

hep-ph↗

arXiv2023

Deblurring Masked Autoencoder is Better Recipe for Ultrasound Image Recognition

Masked autoencoder (MAE) has attracted unprecedented attention and achieves remarkable performance in many vision tasks. It reconstructs random masked image patches (known as proxy task) during pretraining and learns meaningful semantic representations that can be transferred to downstream tasks. However, MAE has not been thoroughly explored in ultrasound imaging. In this work, we investigate the potential of MAE for ultrasound image recognition. Motivated by the unique property of ultrasound imaging in high noise-to-signal ratio, we propose a novel deblurring MAE approach that incorporates deblurring into the proxy task during pretraining. The addition of deblurring facilitates the pretraining to better recover the subtle details presented in the ultrasound images, thus improving the performance of the downstream classification task. Our experimental results demonstrate the effectiveness of our deblurring MAE, achieving state-of-the-art performance in ultrasound image classification. Overall, our work highlights the potential of MAE for ultrasound image recognition and presents a novel approach that incorporates deblurring to further improve its effectiveness.

cs.CV↗

arXiv2023

Scalable generation and detection of on-demand W states in nanophotonic circuits

Quantum physics phenomena, entanglement and coherence, are crucial for quantum information protocols, but understanding these in systems with more than two parts is challenging due to increasing complexity. The W state, a multipartite entangled state, is notable for its robustness and benefits in quantum communication. Here, we generate an 8-mode on-demand single photon W states, using nanowire quantum dots and a silicon nitride photonic chip. We demonstrate a reliable, scalable technique for reconstructing W-state in photonic circuits using Fourier and real-space imaging, supported by the Gerchberg-Saxton phase retrieval algorithm. Additionally, we utilize an entanglement witness to distinguish between mixed and entangled states, thereby affirming the entangled nature of our generated state. The study provides a new imaging approach of assessing multipartite entanglement in W-states, paving the way for further progress in image processing and Fourier-space analysis techniques for complex quantum systems.

quant-ph↗

arXiv2023

Observation of reentrant metal-insulator transition in a random-dimer disordered SSH lattice

The interrelationship between localization, quantum transport, and disorder has remained a fascinating focus in scientific research. Traditionally, it has been widely accepted in the physics community that in one-dimensional systems, as disorder increases, localization intensifies, triggering a metal-insulator transition. However, a recent theoretical investigation [Phys. Rev. Lett. 126, 106803] has revealed that the interplay between dimerization and disorder leads to a reentrant localization transition, constituting a remarkable theoretical advancement in the field. Here, we present the experimental observation of reentrant localization using an experimentally friendly model, a photonic SSH lattice with random-dimer disorder, achieved by incrementally adjusting synthetic potentials. In the presence of correlated on-site potentials, certain eigenstates exhibit extended behavior following the localization transition as the disorder continues to increase. We directly probe the wave function in disordered lattices by exciting specific lattice sites and recording the light distribution. This reentrant phenomenon is further verified by observing an anomalous peak in the normalized participation ratio. Our study enriches the understanding of transport in disordered mediums and accentuates the substantial potential of integrated photonics for the simulation of intricate condensed matter physics phenomena.

cond-mat.dis-nn↗

arXiv2023

Stability through non-shadows

We study families $\mathcal{F}\subseteq 2^{[n]}$ with restricted intersections and prove a conjecture of Snevily in a stronger form for large $n$. We also obtain stability results for Kleitman's isodiametric inequality and families with bounded set-wise differences. Our proofs introduce a new twist to the classical linear algebra method, harnessing the non-shadows of $\mathcal{F}$, which may be of independent interest.

math.CO↗

arXiv2023

A Diffusion Model for Event Skeleton Generation

Event skeleton generation, aiming to induce an event schema skeleton graph with abstracted event nodes and their temporal relations from a set of event instance graphs, is a critical step in the temporal complex event schema induction task. Existing methods effectively address this task from a graph generation perspective but suffer from noise-sensitive and error accumulation, e.g., the inability to correct errors while generating schema. We, therefore, propose a novel Diffusion Event Graph Model~(DEGM) to address these issues. Our DEGM is the first workable diffusion model for event skeleton generation, where the embedding and rounding techniques with a custom edge-based loss are introduced to transform a discrete event graph into learnable latent representation. Furthermore, we propose a denoising training process to maintain the model's robustness. Consequently, DEGM derives the final schema, where error correction is guaranteed by iteratively refining the latent representation during the schema generation process. Experimental results on three IED bombing datasets demonstrate that our DEGM achieves better results than other state-of-the-art baselines. Our code and data are available at https://github.com/zhufq00/EventSkeletonGeneration.

cs.CL↗

arXiv2023

Progressive Learning of 3D Reconstruction Network from 2D GAN Data

This paper presents a method to reconstruct high-quality textured 3D models from single images. Current methods rely on datasets with expensive annotations; multi-view images and their camera parameters. Our method relies on GAN generated multi-view image datasets which have a negligible annotation cost. However, they are not strictly multi-view consistent and sometimes GANs output distorted images. This results in degraded reconstruction qualities. In this work, to overcome these limitations of generated datasets, we have two main contributions which lead us to achieve state-of-the-art results on challenging objects: 1) A robust multi-stage learning scheme that gradually relies more on the models own predictions when calculating losses, 2) A novel adversarial learning pipeline with online pseudo-ground truth generations to achieve fine details. Our work provides a bridge from 2D supervisions of GAN models to 3D reconstruction models and removes the expensive annotation efforts. We show significant improvements over previous methods whether they were trained on GAN generated multi-view images or on real images with expensive annotations. Please visit our web-page for 3D visuals: https://research.nvidia.com/labs/adlr/progressive-3d-learning

cs.CV↗

arXiv2023

Simultaneous CTEQ-TEA extraction of PDFs and SMEFT parameters from jet and $t{\bar t}$ data

Recasting phenomenological Lagrangians in terms of SM effective field theory (SMEFT) provides a valuable means of connecting potential BSM physics at momenta well above the electroweak scale to experimental signatures at lower energies. In this work we jointly fit the Wilson coefficients of SMEFT operators as well as the PDFs in an extension of the CT18 global analysis framework, obtaining self-consistent constraints to possible BSM physics effects. Global fits are boosted with machine-learning techniques in the form of neural networks to ensure efficient scans of the full PDF+SMEFT parameter space. We focus on several operators relevant for top-quark pair and jet production at hadron colliders and obtain constraints on the Wilson coefficients with Lagrange Multiplier scans. We find mild correlations between the extracted Wilson coefficients, PDFs, and other QCD parameters, and see indications that these correlations may become more prominent in future analyses based on data of higher precision. This work serves as a new platform for joint analyses of SM and BSM physics based on the CTEQ-TEA framework.

hep-ph↗