Searcharxiv⌕ Search

arXiv subjects

Gang Chen

Publications and source records attributed to Gang Chen.

At least 253 records · Page 14Linked to original sources
arXiv2023

Photomolecular Effect: Visible Light Interaction with Air-Water Interface

Although water is almost transparent to visible light, we demonstrate that the air-water interface interacts strongly with visible light via what we hypothesize as the photomolecular effect. In this effect, transverse-magnetic polarized photons cleave off water clusters from the air-water interface. We use over 10 different experiments to demonstrate the existence of this effect and its dependence on the wavelength, incident angle and polarization of visible light. We further demonstrate that visible light heats up thin fogs, suggesting that this process can impact weather, climate, and the earth's water cycle. Our study suggests that the photomolecular effect should happen widely in nature, from clouds to fogs, ocean to soil surfaces, and plant transpiration, and can also lead to new applications in energy and clear water.

physics.optics↗
arXiv2023

SPA: A Graph Spectral Alignment Perspective for Domain Adaptation

Unsupervised domain adaptation (UDA) is a pivotal form in machine learning to extend the in-domain model to the distinctive target domains where the data distributions differ. Most prior works focus on capturing the inter-domain transferability but largely overlook rich intra-domain structures, which empirically results in even worse discriminability. In this work, we introduce a novel graph SPectral Alignment (SPA) framework to tackle the tradeoff. The core of our method is briefly condensed as follows: (i)-by casting the DA problem to graph primitives, SPA composes a coarse graph alignment mechanism with a novel spectral regularizer towards aligning the domain graphs in eigenspaces; (ii)-we further develop a fine-grained message propagation module -- upon a novel neighbor-aware self-training mechanism -- in order for enhanced discriminability in the target domain. On standardized benchmarks, the extensive experiments of SPA demonstrate that its performance has surpassed the existing cutting-edge DA methods. Coupled with dense model analysis, we conclude that our approach indeed possesses superior efficacy, robustness, discriminability, and transferability. Code and data are available at: https://github.com/CrownX/SPA.

cs.CV↗
arXiv2023

Continuous Occupancy Mapping in Dynamic Environments Using Particles

Particle-based dynamic occupancy maps were proposed in recent years to model the obstacles in dynamic environments. Current particle-based maps describe the occupancy status in discrete grid form and suffer from the grid size problem, wherein a large grid size is unfavorable for motion planning, while a small grid size lowers efficiency and causes gaps and inconsistencies. To tackle this problem, this paper generalizes the particle-based map into continuous space and builds an efficient 3D egocentric local map. A dual-structure subspace division paradigm, composed of a voxel subspace division and a novel pyramid-like subspace division, is proposed to propagate particles and update the map efficiently with the consideration of occlusions. The occupancy status of an arbitrary point in the map space can then be estimated with the particles' weights. To further enhance the performance of simultaneously modeling static and dynamic obstacles and minimize noise, an initial velocity estimation approach and a mixture model are utilized. Experimental results show that our map can effectively and efficiently model both dynamic obstacles and static obstacles. Compared to the state-of-the-art grid-form particle-based map, our map enables continuous occupancy estimation and substantially improves the performance in different resolutions.

cs.RO↗
arXiv2023

Co-movement Pattern Mining from Videos

Co-movement pattern mining from GPS trajectories has been an intriguing subject in spatial-temporal data mining. In this paper, we extend this research line by migrating the data source from GPS sensors to surveillance cameras, and presenting the first investigation into co-movement pattern mining from videos. We formulate the new problem, re-define the spatial-temporal proximity constraints from cameras deployed in a road network, and theoretically prove its hardness. Due to the lack of readily applicable solutions, we adapt existing techniques and propose two competitive baselines using Apriori-based enumerator and CMC algorithm, respectively. As the principal technical contributions, we introduce a novel index called temporal-cluster suffix tree (TCS-tree), which performs two-level temporal clustering within each camera and constructs a suffix tree from the resulting clusters. Moreover, we present a sequence-ahead pruning framework based on TCS-tree, which allows for the simultaneous leverage of all pattern constraints to filter candidate paths. Finally, to reduce verification cost on the candidate paths, we propose a sliding-window based co-movement pattern enumeration strategy and a hashing-based dominance eliminator, both of which are effective in avoiding redundant operations. We conduct extensive experiments for scalability and effectiveness analysis. Our results validate the efficiency of the proposed index and mining algorithm, which runs remarkably faster than the two baseline methods. Additionally, we construct a video database with 1169 cameras and perform an end-to-end pipeline analysis to study the performance gap between GPS-driven and video-driven methods. Our results demonstrate that the derived patterns from the video-driven approach are similar to those derived from groundtruth trajectories, providing evidence of its effectiveness.

cs.DB↗
arXiv2023

A Simple Text to Video Model via Transformer

We present a general and simple text to video model based on Transformer. Since both text and video are sequential data, we encode both texts and images into the same hidden space, which are further fed into Transformer to capture the temporal consistency and then decoder to generate either text or images. Considering the image signal may become weak in the long sequence, we introduce the U-Net to reconstruct image from its noised version. Specifically, we increase the noise level to the original image in the long sequence, then use the $down$ module from U-Net to encode noised images, which are further input to transformer to predict next clear images. We also add a constraint to promote motion between any generated image pair in the video. We use GPT2 and test our approach on UCF101 dataset and show it can generate promising videos.

cs.CV↗
arXiv2023

Controllable Textual Inversion for Personalized Text-to-Image Generation

The recent large-scale generative modeling has attained unprecedented performance especially in producing high-fidelity images driven by text prompts. Text inversion (TI), alongside the text-to-image model backbones, is proposed as an effective technique in personalizing the generation when the prompts contain user-defined, unseen or long-tail concept tokens. Despite that, we find and show that the deployment of TI remains full of "dark-magics" -- to name a few, the harsh requirement of additional datasets, arduous human efforts in the loop and lack of robustness. In this work, we propose a much-enhanced version of TI, dubbed Controllable Textual Inversion (COTI), in resolving all the aforementioned problems and in turn delivering a robust, data-efficient and easy-to-use framework. The core to COTI is a theoretically-guided loss objective instantiated with a comprehensive and novel weighted scoring mechanism, encapsulated by an active-learning paradigm. The extensive results show that COTI significantly outperforms the prior TI-related approaches with a 26.05 decrease in the FID score and a 23.00% boost in the R-precision.

cs.CV↗
arXiv2023

Upper branch thermal Hall effect in quantum paramagnets

Inspired by the persistent thermal Hall effects at finite temperatures in various quantum paramagnets, we explore the origin of the thermal Hall effects from the perspective of the upper branch parts by invoking the dispersive and twisted crystal field excitations. It is shown that the upper branches of the local energy levels could hybridize and form the dispersive bands. The observation is that, upon the time-reversal symmetry breaking by the magnetic fields, these upper branch bands could acquire a Berry curvature distribution and contribute to the thermal Hall effect in the paramagnetic regime. As a proof of principle, we consider the setting on the kagomé lattice with one ground state singlet and an excited doublet, and show this is indeed possible. We expect this effect to be universal and has no strong connection with the underlying lattice. Although the thermal Hall signal can be contributed from other sources such as phonons and their scattering in the actual materials, we discuss the application to the Mott systems with the large local Hilbert spaces.

cond-mat.str-el↗
arXiv2023

ModelGiF: Gradient Fields for Model Functional Distance

The last decade has witnessed the success of deep learning and the surge of publicly released trained models, which necessitates the quantification of the model functional distance for various purposes. However, quantifying the model functional distance is always challenging due to the opacity in inner workings and the heterogeneity in architectures or tasks. Inspired by the concept of "field" in physics, in this work we introduce Model Gradient Field (abbr. ModelGiF) to extract homogeneous representations from the heterogeneous pre-trained models. Our main assumption underlying ModelGiF is that each pre-trained deep model uniquely determines a ModelGiF over the input space. The distance between models can thus be measured by the similarity between their ModelGiFs. We validate the effectiveness of the proposed ModelGiF with a suite of testbeds, including task relatedness estimation, intellectual property protection, and model unlearning verification. Experimental results demonstrate the versatility of the proposed ModelGiF on these tasks, with significantly superiority performance to state-of-the-art competitors. Codes are available at https://github.com/zju-vipa/modelgif.

cs.LG↗
arXiv2023

Enhancement of Quantum Sensing in a Cavity Optomechanical System around Quantum Critical Point

The precision of quantum sensing could be improved by exploiting quantum phase transitions, where the physical quantity tends to diverge when the system approaches the quantum critical point. This critical enhancement phenomenon has been applied to the quantum Rabi model in a dynamic framework, showing a promising sensing enhancement without the need for complex initial state preparation. In this work, we present a quantum phase transition in the coupling cavity-mechanical oscillator system when the coupling strength crosses a critical point, determined by the effective detuning of cavity and frequency of mechanical mode. By utilizing this critical phenomenon, we obtain a prominent enhancement of quantum sensing, such as the position and momentum of the mechanical oscillator. This result provides an alternative method to enhance the quantum sensing of some physical quantities, such as mass, charge, and weak force, in a large mass system.

quant-ph↗
arXiv2023

An introduction to the parton and hadron cascade model PACIAE 3.0

We introduce a parton and hadron cascade model PACIAE 3.0 based on PYTHIA 6.428 and the PACIAE 2.2 program series. The simulation framework of C-, B-, and A-loops are designed for the high energy ($\sqrt{s_{NN}}\geq 3$ GeV) and low energy ($\sqrt{s_{NN}}<3$ GeV) nuclear collisions, respectively, in PACIAE 3.0. In the C-loop simulation, the parton-parton inelastic scattering processes are added in the partonic rescattering process. The single string structure and multiple string interaction mechanism have been introduced investigating the strangeness enhancement in C- and B-loop. An improved mapping relation between the centrality percentage definition and the impact parameter definition is proposed responding the observation of $b_{max}\approx 20$ fm from ALICE, ATLAS, and CMS collaborations. We have extensively modified the phenomenological coalescence hadronization model. The PACIAE 3.0 model simulated results of particle yield, transverse momentum distribution, and rapidity distribution well reproduce, respectively, the experimental data measured at FOPI, E895, RHIC, and LHC energies.

hep-ph↗
arXiv2023

Assessing Hidden Risks of LLMs: An Empirical Study on Robustness, Consistency, and Credibility

The recent popularity of large language models (LLMs) has brought a significant impact to boundless fields, particularly through their open-ended ecosystem such as the APIs, open-sourced models, and plugins. However, with their widespread deployment, there is a general lack of research that thoroughly discusses and analyzes the potential risks concealed. In that case, we intend to conduct a preliminary but pioneering study covering the robustness, consistency, and credibility of LLMs systems. With most of the related literature in the era of LLM uncharted, we propose an automated workflow that copes with an upscaled number of queries/responses. Overall, we conduct over a million queries to the mainstream LLMs including ChatGPT, LLaMA, and OPT. Core to our workflow consists of a data primitive, followed by an automated interpreter that evaluates these LLMs under different adversarial metrical systems. As a result, we draw several, and perhaps unfortunate, conclusions that are quite uncommon from this trendy community. Briefly, they are: (i)-the minor but inevitable error occurrence in the user-generated query input may, by chance, cause the LLM to respond unexpectedly; (ii)-LLMs possess poor consistency when processing semantically similar query input. In addition, as a side finding, we find that ChatGPT is still capable to yield the correct answer even when the input is polluted at an extreme level. While this phenomenon demonstrates the powerful memorization of the LLMs, it raises serious concerns about using such data for LLM-involved evaluation in academic development. To deal with it, we propose a novel index associated with a dataset that roughly decides the feasibility of using such data for LLM-involved evaluation. Extensive empirical studies are tagged to support the aforementioned claims.

cs.LG↗
arXiv2023

Distinguishing thermodynamics and spectroscopy for octupolar U(1) spin liquid of Ce-pyrochlores

Inspired by the progress on the spin liquid candidates Ce$_2$Sn$_2$O$_7$ and Ce$_2$Zr$_2$O$_7$ where the Ce ions carry the dipole-octupole doublets, we analyze the distinction between the thermodynamic and spectroscopic measurements for the octupolar U(1) spin liquid. Due to the peculiar properties of octupolar U(1) spin liquid and the selective Zeeman coupling, both uniform susceptibility and the spectroscopic measurements only probe the spinon excitations, while the specific heat measures all excitations. After clarifying the contents in each measurement, we observe that, thermodynamics carries the information of single spinon, while the spectroscopy is associated with spinon pairs. Such distinction immediately leads to the multiplicity relation of the excitation gaps from different measurements that include the uniform susceptibility, Knight shift and $1/T_1$ spin lattice relaxation of NMR and $μ$SR, and the inelastic neutron scattering measurements. We hope this work provides a useful recipe for the identification of fractionalization in the octupolar U(1) spin liquid and other spin liquids.

cond-mat.str-el↗
arXiv2023

Coulombic antiferromagnet in spin-1/2 pyrochlores with dipole-octupole doublets

Exotic state of matter could coexist with conventional orders such that the gauge fields and fractionalized excitations could prevail in a seemingly ordered state. We explore the Coulombic antiferromagnet in the spin-1/2 pyrochlores with dipole-octupole doublets. This fractionalized antiferromagnetic state carries both antiferromagnetic order and emergent quantum electrodynamics with the gapless gauge photon and fractionalized quasiparticles. We explain the characteristic physical properties including the thermodynamics and the dynamics of this exotic state. This includes for example the anomalously large $T^3$ specific heat and the broad spinon continuum in the inelastic neutron scattering. We discuss the experimental and material's relevance and expect this work to inspire interests in the search of exotic physics among the ordered magnets.

cond-mat.str-el↗
arXiv2023

Quadrupole moments and their interactions in the triangular lattice antiferromagnet FeI$_2$

Motivated by the recent experiments on the triangular lattice antiferromagnet FeI$_2$, we consider the presence of the quadrupole moments and their interactions between the spin-orbital-entangled local moments. In addition to the anisotropic pairwise interaction between the local moments, the interaction between the quadrupole moment arises from the interaction of the orbital occupation configurations. This is argued primarily from the time reversal symmetry and the different orbital configurations via their exchange paths. We discuss the implication of these interactions and expect this result to be complementary to the existing works in this system and other systems alike.

cond-mat.str-el↗
arXiv2023

Minimizing the number of edges in $(C_4, K_{1,k})$-co-critical graphs

Given graphs $H_1, H_2$, a {red, blue}-coloring of the edges of a graph $G$ is a critical coloring if $G$ has neither a red $H_1$ nor a blue $ H_2$. A non-complete graph $G$ is $(H_1, H_2)$-co-critical if $G$ admits a critical coloring, but $G+e$ has no critical coloring for every edge $e$ in the complement of $G$. Motivated by a conjecture of Hanson and Toft from 1987, we study the minimum number of edges over all $(C_4, K_{1,k})$-co-critical graphs on $n$ vertices. We show that for all $k \ge 2 $ and $ n \ge k +\lfloor \sqrt {k-1} \rfloor +2$, if $G$ is a $(C_4,K_{1,k})$-co-critical graph on $n$ vertices, then \[e(G) \ge \frac{(k+2)n}2-3- \frac{(k-1)(k+ \lfloor \sqrt {k-2}\rfloor)}2.\] Moreover, this linear bound is asymptotically best possible for all $k\ge3$ and $n\ge3k+4$. It is worth noting that our constructions for the case when $ k$ is even have at least three different critical colorings. For $k=2$, we obtain the sharp bound for the minimum number of edges of $(C_4, K_{1,2})$-co-critical graphs on $n\ge5 $ vertices by showing that all such graphs have at least $2n-3$ edges. Our proofs rely on the structural properties of $(C_4,K_{1,k})$-co-critical graphs and a result of Ollmann on the minimum number of edges of $C_4$-saturated graphs.

math.CO↗
arXiv2023

TableGPT: Towards Unifying Tables, Nature Language and Commands into One GPT

Tables are prevalent in real-world databases, requiring significant time and effort for humans to analyze and manipulate. The advancements in large language models (LLMs) have made it possible to interact with tables using natural language input, bringing this capability closer to reality. In this paper, we present TableGPT, a unified fine-tuned framework that enables LLMs to understand and operate on tables using external functional commands. It introduces the capability to seamlessly interact with tables, enabling a wide range of functionalities such as question answering, data manipulation (e.g., insert, delete, query, and modify operations), data visualization, analysis report generation, and automated prediction. TableGPT aims to provide convenience and accessibility to users by empowering them to effortlessly leverage tabular data. At the core of TableGPT lies the novel concept of global tabular representations, which empowers LLMs to gain a comprehensive understanding of the entire table beyond meta-information. By jointly training LLMs on both table and text modalities, TableGPT achieves a deep understanding of tabular data and the ability to perform complex operations on tables through chain-of-command instructions. Importantly, TableGPT offers the advantage of being a self-contained system rather than relying on external API interfaces. Moreover, it supports efficient data process flow, query rejection (when appropriate) and private deployment, enabling faster domain data fine-tuning and ensuring data privacy, which enhances the framework's adaptability to specific use cases.

cs.AI↗
arXiv2023

On Paradoxical Phenomena During Evaporation and Condensation between Two Parallel Plates

Kinetic theory has long predicted that temperature inversion may happen in the vapor-phase for evaporation and condensation between two parallel plates, i.e., the vapor temperature at the condensation interface is higher than that at the evaporation interface. However, past studies have neglected transport in the liquid phases, which usually determine the evaporation and condensation rates. This disconnect has limited the acceptance of the kinetic theory in practical heat transfer models. In this paper, we combine interfacial conditions for mass and heat fluxes with continuum descriptions in the bulk regions of the vapor and the liquid phases to obtain a complete picture for the classical problem of evaporation and condensation between two parallel plates. The criterion for temperature inversion is rederived analytically. We also prove that the temperature jump at each interface is in the same direction as externally applied temperature difference, i.e., liquid surface is at a higher temperature than its adjacent vapor on the evaporating interface and at a lower temperature than its adjacent vapor on the condensing interface. We explain the interfacial temperature jump and temperature inversion using the interfacial cooling and heating processes, and we predict that this process can lead to a vapor phase temperature much lower than the lowest wall temperatures and much higher than the highest wall temperature imposed. When the latent heat of evaporation is small, we found that evaporation can happen at the low temperature side while condensation occur at the high temperature side, opposing the temperature gradient.

physics.flu-dyn↗
arXiv2023

ProMix: Combating Label Noise via Maximizing Clean Sample Utility

Learning with Noisy Labels (LNL) has become an appealing topic, as imperfectly annotated data are relatively cheaper to obtain. Recent state-of-the-art approaches employ specific selection mechanisms to separate clean and noisy samples and then apply Semi-Supervised Learning (SSL) techniques for improved performance. However, the selection step mostly provides a medium-sized and decent-enough clean subset, which overlooks a rich set of clean samples. To fulfill this, we propose a novel LNL framework ProMix that attempts to maximize the utility of clean samples for boosted performance. Key to our method, we propose a matched high confidence selection technique that selects those examples with high confidence scores and matched predictions with given labels to dynamically expand a base clean sample set. To overcome the potential side effect of excessive clean set selection procedure, we further devise a novel SSL framework that is able to train balanced and unbiased classifiers on the separated clean and noisy samples. Extensive experiments demonstrate that ProMix significantly advances the current state-of-the-art results on multiple benchmarks with different types and levels of noise. It achieves an average improvement of 2.48\% on the CIFAR-N dataset. The code is available at https://github.com/Justherozen/ProMix

cs.LG↗