Searcharxiv⌕ Search

arXiv subjects

Xing Wang

Publications and source records attributed to Xing Wang.

At least 127 records · Page 7Linked to original sources
arXiv2023

Leveraging Word Guessing Games to Assess the Intelligence of Large Language Models

The automatic evaluation of LLM-based agent intelligence is critical in developing advanced LLM-based agents. Although considerable effort has been devoted to developing human-annotated evaluation datasets, such as AlpacaEval, existing techniques are costly, time-consuming, and lack adaptability. In this paper, inspired by the popular language game ``Who is Spy'', we propose to use the word guessing game to assess the intelligence performance of LLMs. Given a word, the LLM is asked to describe the word and determine its identity (spy or not) based on its and other players' descriptions. Ideally, an advanced agent should possess the ability to accurately describe a given word using an aggressive description while concurrently maximizing confusion in the conservative description, enhancing its participation in the game. To this end, we first develop DEEP to evaluate LLMs' expression and disguising abilities. DEEP requires LLM to describe a word in aggressive and conservative modes. We then introduce SpyGame, an interactive multi-agent framework designed to assess LLMs' intelligence through participation in a competitive language-based board game. Incorporating multi-agent interaction, SpyGame requires the target LLM to possess linguistic skills and strategic thinking, providing a more comprehensive evaluation of LLMs' human-like cognitive abilities and adaptability in complex communication situations. The proposed evaluation framework is very easy to implement. We collected words from multiple sources, domains, and languages and used the proposed evaluation framework to conduct experiments. Extensive experiments demonstrate that the proposed DEEP and SpyGame effectively evaluate the capabilities of various LLMs, capturing their ability to adapt to novel situations and engage in strategic communication.

cs.CL↗
arXiv2023

Is ChatGPT A Good Translator? Yes With GPT-4 As The Engine

This report provides a preliminary evaluation of ChatGPT for machine translation, including translation prompt, multilingual translation, and translation robustness. We adopt the prompts advised by ChatGPT to trigger its translation ability and find that the candidate prompts generally work well with minor performance differences. By evaluating on a number of benchmark test sets, we find that ChatGPT performs competitively with commercial translation products (e.g., Google Translate) on high-resource European languages but lags behind significantly on low-resource or distant languages. As for the translation robustness, ChatGPT does not perform as well as the commercial systems on biomedical abstracts or Reddit comments but exhibits good results on spoken language. Further, we explore an interesting strategy named $\mathbf{pivot~prompting}$ for distant languages, which asks ChatGPT to translate the source sentence into a high-resource pivot language before into the target language, improving the translation performance noticeably. With the launch of the GPT-4 engine, the translation performance of ChatGPT is significantly boosted, becoming comparable to commercial translation products, even for distant languages. Human analysis on Google Translate and ChatGPT suggests that ChatGPT with GPT-3.5 tends to generate more hallucinations and mis-translation errors while that with GPT-4 makes the least errors. In other words, ChatGPT has already become a good translator. Please refer to our Github project for more details: https://github.com/wxjiao/Is-ChatGPT-A-Good-Translator

cs.CL↗
arXiv2023

ParroT: Translating during Chat using Large Language Models tuned with Human Translation and Feedback

Large language models (LLMs) like ChatGPT have exhibited remarkable abilities on a wide range of natural language processing~(NLP) tasks, including various machine translation abilities accomplished during chat. However, these models are only accessible through restricted APIs, which creates barriers to new research and advancements in the field. Therefore, we propose ParroT, a framework to enhance and regulate the translation abilities during chat based on open-source LLMs (e.g., LLaMA), human-written translation and feedback data. Specifically, ParroT reformulates translation data into the instruction-following style, and introduces a "$\mathbf{Hint}$" field for incorporating extra requirements to regulate the translation process. Accordingly, we propose three instruction types for finetuning ParroT models, including translation instruction, contrastive instruction, and error-guided instruction. Experiments on Flores subsets and WMT22 test sets suggest that translation instruction improves the translation performance of vanilla LLMs significantly while error-guided instruction can lead to further improvement, which demonstrates the importance of learning from low-quality translations annotated by humans. We also demonstrate the potential of automatic evaluation tools in providing quality information of translations, when constructing error-guided instructions for directions that lack human annotation data. Please refer to our Github project for more implementation details: https://github.com/wxjiao/ParroT

cs.CL↗
arXiv2023

Charge equilibration of Laser-accelerated Carbon Ions in Foam Target

The charge equilibration of laser-accelerated carbon ion beams in 2 mg/cm3 foam target was investigated experimentally. The ions were generated through target normal sheath acceleration mechanism in laser-foil interaction scheme. This allows to get the equilibrium charge state in wide energy range near Bragg peak within a single shot. By using foam, the charge equilibration measurement in density regime between gas and solid state was firstly reached out experimentally. It was found that the theoretical predictions with tabulated cross section data for gas target greatly underestimated the charge states. The experimental data are in close agreement with both semi-empirical formula as well as rate equation predictions based on ion-solid interactions. The important role of target density effects that increase the ionization probability and decrease the electron capture probability through frequent multi-collisions in foam are demonstrated. The double electron processes are shown to have little influence on the average charge states. The findings are essential for high energy density physics research where the foams are widely used, and have impacts on a broad range of applications in medical, biological and material fields. The method also provides a new approach to investigate the interaction mechanism of swift heavy ions in matter by taking advantage of the laser-accelerated short-pulse wide-energy range ions.

physics.atom-ph↗
arXiv2023

Feynman integrals, geometries and differential equations

In this talk we discuss the construction of a basis of master integrals for the family of the $l$-loop equal-mass banana integrals, such that the differential equation is in an $\varepsilon$-factorised form. As the $l$-loop banana integral is related to a Calabi-Yau $(l-1)$-fold, this extends the examples where an $\varepsilon$-factorised form has been found from Feynman integrals related to curves (of genus zero and one) to Feynman integrals related to higher-dimensional varieties.

hep-th↗
arXiv2023

$\varepsilon$-forms for non-planar triangles with elliptic curves at two loops

In this talk, we discuss how to generalize ideas developed for Banana integrals to two two-loop non-planar triangle Feynman integrals involving elliptic curves, which have non-trivial sub-sectors and whose Picard-Fuchs operators share less symmetry than Banana integrals, to obtain the canonical differential equations and to solve them with suitable boundary conditions.

hep-th↗
arXiv2023

$\varepsilon$-factorized differential equations for two-loop non-planar triangle Feynman integrals with elliptic curves

In this paper, we investigate two-loop non-planar triangle Feynman integrals involving elliptic curves. In contrast to the Sunrise and Banana integral families, the triangle families involve non-trivial sub-sectors. We show that the methodology developed in the context of Banana integrals can also be extended to these cases and obtain $\varepsilon$-factorized differential equations for all sectors. The letters are combinations of modular forms on the corresponding elliptic curves and algebraic functions arising from the sub-sectors. With uniform transcendental boundary conditions, we express our results in terms of iterated integrals order-by-order in the dimensional regulator, which can be evaluated efficiently. Our method can be straightforwardly generalized to other elliptic integral families and have important applications to precision physics at current and future high-energy colliders.

hep-th↗
arXiv2023

Vector-Like Top Quark Production via an Electroweak Dipole Moment at a Muon Collider

Vectorial partners of the Standard Model quarks and leptons are predicted in many dynamical models of electroweak symmetry breaking. The most easily accessible of these new particles, either due to mass or couplings, are typically expected to be the partners of the third-generation fermions. It is therefore essential to explore the signatures of these particles at future high-energy colliders. We study the potential of a high-energy muon collider to singly produce a vector-like top-quark partner via an electroweak dipole moment operator, such an operator being typical of composite constructions beyond the Standard Model. We use a phenomenological model for third-generation quarks and their partners that satisfies an extended custodial symmetry. This automatically protects the $W$-boson and $Z$-boson masses from receiving large electroweak corrections, and it allows the model to be viable given current electroweak data. We demonstrate that cross sections associated with dipole-induced vector-like quark production can easily exceed those inherent to more conventional single-production modes via ordinary electroweak couplings. We then explore the associated phenomenology, and we show that at least one (and often more than one) of the extra vector-like states can be studied at high-energy muon colliders. Typical accessible masses are found to range up to close to the kinematic production threshold, when the vector-like partners are produced in combination with an ordinary top quark.

hep-ph↗
arXiv2023

SepViT: Separable Vision Transformer

Vision Transformers have witnessed prevailing success in a series of vision tasks. However, these Transformers often rely on extensive computational costs to achieve high performance, which is burdensome to deploy on resource-constrained devices. To alleviate this issue, we draw lessons from depthwise separable convolution and imitate its ideology to design an efficient Transformer backbone, i.e., Separable Vision Transformer, abbreviated as SepViT. SepViT helps to carry out the local-global information interaction within and among the windows in sequential order via a depthwise separable self-attention. The novel window token embedding and grouped self-attention are employed to compute the attention relationship among windows with negligible cost and establish long-range visual interactions across multiple windows, respectively. Extensive experiments on general-purpose vision benchmarks demonstrate that SepViT can achieve a state-of-the-art trade-off between performance and latency. Among them, SepViT achieves 84.2% top-1 accuracy on ImageNet-1K classification while decreasing the latency by 40%, compared to the ones with similar accuracy (e.g., CSWin). Furthermore, SepViT achieves 51.0% mIoU on ADE20K semantic segmentation task, 47.9 AP on the RetinaNet-based COCO detection task, 49.4 box AP and 44.6 mask AP on Mask R-CNN-based COCO object detection and instance segmentation tasks.

cs.CV↗
arXiv2023

MPPN: Multi-Resolution Periodic Pattern Network For Long-Term Time Series Forecasting

Long-term time series forecasting plays an important role in various real-world scenarios. Recent deep learning methods for long-term series forecasting tend to capture the intricate patterns of time series by decomposition-based or sampling-based methods. However, most of the extracted patterns may include unpredictable noise and lack good interpretability. Moreover, the multivariate series forecasting methods usually ignore the individual characteristics of each variate, which may affecting the prediction accuracy. To capture the intrinsic patterns of time series, we propose a novel deep learning network architecture, named Multi-resolution Periodic Pattern Network (MPPN), for long-term series forecasting. We first construct context-aware multi-resolution semantic units of time series and employ multi-periodic pattern mining to capture the key patterns of time series. Then, we propose a channel adaptive module to capture the perceptions of multivariate towards different patterns. In addition, we present an entropy-based method for evaluating the predictability of time series and providing an upper bound on the prediction accuracy before forecasting. Our experimental evaluation on nine real-world benchmarks demonstrated that MPPN significantly outperforms the state-of-the-art Transformer-based, decomposition-based and sampling-based methods for long-term series forecasting.

cs.LG↗
arXiv2023

A Guide to Diagnosing Colored Resonances at Hadron Colliders

We present a comprehensive study on how to distinguish the properties of heavy dijet resonances at hadron colliders. A variety of spins, chiral couplings, charges, and QCD color representations are considered. Distinguishing the different color representations is particularly difficult at hadron colliders. To determine the QCD color structure, we consider a third jet radiated in a resonant dijet event. We show that the relative rates of three-jet versus two-jet processes are sensitive to the color representation of the resonance. We also show analytically that the antennae radiation pattern of soft radiation depends on the color structure of dijet events and develops an observable that is sensitive to the antennae patterns. Finally, we exploit a Convolutional Neural Network with Machine Learning techniques to differentiate the radiation patterns from different colored resonances and find encouraging results to discriminate them. We demonstrate our results numerically at a 14 TeV LHC, and the methodology presented here should be applicable to other future hadron colliders.

hep-ph↗
arXiv2023

Energy loss enhancement of very intense proton beams in dense matter due to the beam-density effect

Thoroughly understanding the transport and energy loss of intense ion beams in dense matter is essential for high-energy-density physics and inertial confinement fusion. Here, we report a stopping power experiment with a high-intensity laser-driven proton beam in cold, dense matter. The measured energy loss is one order of magnitude higher than the expectation of individual particle stopping models. We attribute this finding to the proximity of beam ions to each other, which is usually insignificant for relatively-low-current beams from classical accelerators. The ionization of the cold target by the intense ion beam is important for the stopping power calculation and has been considered using proper ionization cross section data. Final theoretical values agree well with the experimental results. Additionally, we extend the stopping power calculation for intense ion beams to plasma scenario based on Ohm's law. Both the proximity- and the Ohmic effect can enhance the energy loss of intense beams in dense matter, which are also summarized as the beam-density effect. This finding is useful for the stopping power estimation of intense beams and significant to fast ignition fusion driven by intense ion beams.

physics.plasm-ph↗
arXiv2023

Conversion of La$_2$Ti$_2$O$_7$ to LaTiO$_2$N via Ammonolysis: An ab-initio Investigation

Perovskite oxynitrides are, due to their reduced band gap compared to oxides, promising materials for photocatalytic applications. They are most commonly synthesized from {110} layered Carpy-Galy (A$_2$B$_2$O$_7$}) perovskites via thermal ammonolysis, i.e. the exposure to a flow of ammonia at elevated temperature. The conversion of the layered oxide to the non-layered oxynitride must involve a complex combination of nitrogen incorporation, oxygen removal and ultimately structural transition by elimination of the interlayer shear plane. Despite the process being commonly used, little is known about the microscopic mechanisms and hence factors that could ease the conversion. Here we aim to derive such insights via density functional theory calculations of the defect chemistry of the oxide and the oxynitride as well as the oxide's surface chemistry. Our results point to the crucial role of surface oxygen vacancies in forming clusters of NH$_3$ decomposition products and in incorporating N, most favorably substitutionally at the anion site. N then spontaneously diffuses away from the surface, more easily parallel to the surface and in interlayer regions, while diffusion perpendicular to the interlayer plane is somewhat slower. Once incorporation and diffusion lead to a local N concentration of about 70% of the stoichiometric oxynitride composition, the nitridated oxide spontaneously transforms to a nitrogen-deficient oxynitride.

cond-mat.mtrl-sci↗
arXiv2023

Transport of intense ion beams in plasmas: collimation and energy-loss reduction

We compare the transport properties of a well-characterized hydrogen plasma for low and high current ion beams. The energy-loss of low current beams can be well understood, within the framework of current stopping power models. However, for high current proton beams, significant energy-loss reduction and collimation is observed in the experiment. We have developed a new particle-in-cell code, which includes both collective electromagnetic effects and collisional interactions. Our simulations indicate that resistive magnetic fields, induced by the transport of an intense proton beam, act to collimate the proton beam and simultaneously deplete the local plasma density along the beam path. This in turn causes the energy-loss reduction detected in the experiment.

physics.plasm-ph↗
arXiv2023

Bananas of equal mass: any loop, any order in the dimensional regularisation parameter

We describe a systematic approach to cast the differential equation for the $l$-loop equal mass banana integral into an $\varepsilon$-factorised form. With the known boundary value at a specific point we obtain systematically the term of order $j$ in the expansion in the dimensional regularisation parameter $\varepsilon$ for any loop $l$. The approach is based on properties of Calabi-Yau operators, and in particular on self-duality.

hep-th↗
arXiv2023

Adaptive Hybrid Spatial-Temporal Graph Neural Network for Cellular Traffic Prediction

Cellular traffic prediction is an indispensable part for intelligent telecommunication networks. Nevertheless, due to the frequent user mobility and complex network scheduling mechanisms, cellular traffic often inherits complicated spatial-temporal patterns, making the prediction incredibly challenging. Although recent advanced algorithms such as graph-based prediction approaches have been proposed, they frequently model spatial dependencies based on static or dynamic graphs and neglect the coexisting multiple spatial correlations induced by traffic generation. Meanwhile, some works lack the consideration of the diverse cellular traffic patterns, result in suboptimal prediction results. In this paper, we propose a novel deep learning network architecture, Adaptive Hybrid Spatial-Temporal Graph Neural Network (AHSTGNN), to tackle the cellular traffic prediction problem. First, we apply adaptive hybrid graph learning to learn the compound spatial correlations among cell towers. Second, we implement a Temporal Convolution Module with multi-periodic temporal data input to capture the nonlinear temporal dependencies. In addition, we introduce an extra Spatial-Temporal Adaptive Module to conquer the heterogeneity lying in cell towers. Our experiments on two real-world cellular traffic datasets show AHSTGNN outperforms the state-of-the-art by a significant margin, illustrating the superior scalability of our method for spatial-temporal cellular traffic prediction.

cs.NI↗
arXiv2023

The misuse of the nonlinear field normalization method: Nonlinear field normalization citation counts at the paper level cannot be added or averaged

There is a very important problem that has not attracted sufficient attention in academia, i.e., nonlinear field normalization citation counts at the paper level obtained using nonlinear field normalization methods cannot be added or averaged. Unfortunately, there are many cases adding or averaging the nonlinear normalized citation counts of individual papers that can be found in the academic literature, indicating that nonlinear field normalization methods have long been misused in academia. In this paper, we performed the following two research works. First, we analyzed why the nonlinear normalized citation counts of individual papers cannot be added or averaged from the perspective of theoretical analysis in mathematics: we provide mathematical proofs for the crucial steps of the analysis. Second, we systematically classified the existing main field normalization methods into linear and nonlinear field normalization methods. The above two research works provide a theoretical basis for the proper use of field normalization methods in the future, avoiding the continued misuse of nonlinear data. Furthermore, because our mathematical proof is applicable to all nonlinear data in the entire real number domain, our research works are also meaningful for the whole field of data and information science.

cs.DL↗
arXiv2023

Scaling Back-Translation with Domain Text Generation for Sign Language Gloss Translation

Sign language gloss translation aims to translate the sign glosses into spoken language texts, which is challenging due to the scarcity of labeled gloss-text parallel data. Back translation (BT), which generates pseudo-parallel data by translating in-domain spoken language texts into sign glosses, has been applied to alleviate the data scarcity problem. However, the lack of large-scale high-quality domain spoken language text data limits the effect of BT. In this paper, to overcome the limitation, we propose a Prompt based domain text Generation (PGEN) approach to produce the large-scale in-domain spoken language text data. Specifically, PGEN randomly concatenates sentences from the original in-domain spoken language text data as prompts to induce a pre-trained language model (i.e., GPT-2) to generate spoken language texts in a similar style. Experimental results on three benchmarks of sign language gloss translation in varied languages demonstrate that BT with spoken language texts generated by PGEN significantly outperforms the compared methods. In addition, as the scale of spoken language texts generated by PGEN increases, the BT technique can achieve further improvements, demonstrating the effectiveness of our approach. We release the code and data for facilitating future research in this field.

cs.CL↗