arXiv subjects

Jiaxun Li

Publications and source records attributed to Jiaxun Li.

12 recordsLinked to original sources

arXiv2026

Characterizing and spectrally modeling embedded FUor eruptions in the near-infrared

Context. Episodic accretion in young stellar objects (YSOs) is thought to play a critical role in addressing the "luminosity problem" associated with star formation. However, optical surveys tend to bias against sources that are heavily obscured. Infrared time-domain surveys, such as unTimely WISE, facilitate the identification of such sources within the dense star formation regions of our Galaxy. Aims. We aim to systematically identify and characterize FUor outbursts in infrared-selected YSOs using high-resolution spectroscopy and detailed disk modeling. Methods. We conducted follow-up high-resolution spectroscopy with Gemini South/IGRINS for four FUor candidates discovered in infrared time-domain surveys. Using a combination of photometric and spectroscopic observations, we constructed spectral energy distributions and fit them with a disk model that incorporates an actively accreting inner disk together with a passively irradiated outer disk. Results. All objects show CO and H$_2$O absorption bands at 2.3$μ$m, and their positions in the Na + Ca versus CO equivalent width diagram further corroborate their classification as FUors. The best-fitting model spectra closely match both the observed spectral features and the overall continuum, providing additional confirmation of the FUor classification. The best-fit models reveal high extinction values ($A_V$ = 10-20 mag), with $M_*\dot{M}$ comparable to those of classical FUors such as FU Orionis. Among 18 sources initially selected via infrared light curves, $6-$7 out of 8 with available spectra exhibit FUor characteristics, implying a high selection efficiency.

astro-ph.SR↗

arXiv2026

On Generation in Metric Spaces

We study generation in separable metric instance spaces. We extend the language generation framework from Kleinberg and Mullainathan [2024] beyond countable domains by defining novelty through metric separation and allowing asymmetric novelty parameters for the adversary and the generator. We introduce the $(\varepsilon,\varepsilon')$-closure dimension, a scale-sensitive analogue of closure dimension, which yields characterizations of uniform and non-uniform generatability and a sufficient condition for generation in the limit. Along the way, we identify a sharp geometric contrast. Namely, in doubling spaces, including all finite-dimensional normed spaces, generatability is stable across novelty scales and invariant under equivalent metrics. In general metric spaces, however, generatability can be highly scale-sensitive and metric-dependent; even in the natural infinite-dimensional Hilbert space $\ell^2$, all notions of generation may fail abruptly as the novelty parameters vary.

stat.ML↗

arXiv2025

Mutual Information Guided Backdoor Mitigation for Pre-trained Encoders

Self-supervised learning (SSL) is increasingly attractive for pre-training encoders without requiring labeled data. Downstream tasks built on top of those pre-trained encoders can achieve nearly state-of-the-art performance. The pre-trained encoders by SSL, however, are vulnerable to backdoor attacks as demonstrated by existing studies. Numerous backdoor mitigation techniques are designed for downstream task models. However, their effectiveness is impaired and limited when adapted to pre-trained encoders, due to the lack of label information when pre-training. To address backdoor attacks against pre-trained encoders, in this paper, we innovatively propose a mutual information guided backdoor mitigation technique, named MIMIC. MIMIC treats the potentially backdoored encoder as the teacher net and employs knowledge distillation to distill a clean student encoder from the teacher net. Different from existing knowledge distillation approaches, MIMIC initializes the student with random weights, inheriting no backdoors from teacher nets. Then MIMIC leverages mutual information between each layer and extracted features to locate where benign knowledge lies in the teacher net, with which distillation is deployed to clone clean features from teacher to student. We craft the distillation loss with two aspects, including clone loss and attention loss, aiming to mitigate backdoors and maintain encoder performance at the same time. Our evaluation conducted on two backdoor attacks in SSL demonstrates that MIMIC can significantly reduce the attack success rate by only utilizing <5% of clean data, surpassing seven state-of-the-art backdoor mitigation techniques.

cs.LG↗

arXiv2025

Exact and Conservative Inference for the Average Treatment Effect in Stratified Experiments with Binary Outcomes

We extend methods for finite-sample inference about the average treatment effect (ATE) in randomized experiments with binary outcomes to accommodate stratification (blocking). We present three valid methods that differ in their computational and statistical efficiency. The first method constructs conservative, Bonferroni-adjusted confidence intervals separately for the mean response in the treatment and control groups in each stratum, then takes appropriate weighted differences of their endpoints to find a confidence interval for the ATE. The second method inverts permutation tests for the overall ATE, maximizing the $P$-value over all ways a given ATE can be attained. The third method applies permutation tests for the ATE in separate strata, then combines those tests to form a confidence interval for the overall ATE. We compare the statistical and computational performance of the methods using simulations and a case study. The second approach is most efficient statistically in the simulations, but a naive implementation requires O(Π_{k=1}^{K} n_{k}^{4}) permutation tests, the highest computational burden among the three methods. That computational burden can be reduced to O(\sum_{k=1}^K n_k \timesΠ_{k=1}^{K} n_{k}^{2}) if all strata are balanced and to O(Π_{k=1}^{K} n_{k}^{3}) otherwise.

stat.ME↗

arXiv2024

Generation through the lens of learning theory

We study generation through the lens of statistical learning theory. First, we abstract and formalize the results of Gold [1967], Angluin [1979], Angluin [1980] and Kleinberg and Mullainathan [2024] in terms of a binary hypothesis class defined over an abstract example space. Then, we extend the notion of "generation" from Kleinberg and Mullainathan [2024] to two new settings, we call "uniform" and "non-uniform" generation, and provide a characterization of which hypothesis classes are uniformly and non-uniformly generatable. As is standard in learning theory, our characterizations are in terms of the finiteness of a new combinatorial dimension termed the Closure dimension. By doing so, we are able to compare generatability with predictability (captured via PAC and online learnability) and show that these two properties of hypothesis classes are incompatible -- there are classes that are generatable but not predictable and vice versa. Finally, we extend our results to capture prompted generation and give a complete characterization of which classes are prompt generatable, generalizing some of the work by Kleinberg and Mullainathan [2024].

cs.LG↗

arXiv2024

HGC: A hybrid method combining gravity model and cycle structure for identifying influential spreaders in complex networks

Identifying influential spreaders in complex networks is a critical challenge in network science, with broad applications in disease control, information dissemination, and influence analysis in social networks. The gravity model, a distinctive approach for identifying influential spreaders, has attracted significant attention due to its ability to integrate node influence and the distance between nodes. However, the law of gravity is symmetric, whereas the influence between different nodes is asymmetric. Existing gravity model-based methods commonly rely on the topological distance as a metric to measure the distance between nodes. Such reliance neglects the strength or frequency of connections between nodes, resulting in symmetric influence values between node pairs, which ultimately leads to an inaccurate assessment of node influence. Moreover, these methods often overlook cycle structures within networks, which provide redundant pathways for nodes and contribute significantly to the overall connectivity and stability of the network. In this paper, we propose a hybrid method called HGC, which integrates the gravity model with effective distance and incorporates cycle structure to address the issues above. Effective distance, derived from probabilities, measures the distance between a source node and others by considering its connectivity, providing a more accurate reflection of actual relationships between nodes. To evaluate the accuracy and effectiveness of the proposed method, we conducted several experiments on eight real-world networks based on the Susceptible-Infected-Recovered model. The results demonstrate that HGC outperforms seven compared methods in accurately identifying influential nodes.

cs.CE↗

arXiv2024

A population of mid-infrared large-amplitude variable young stellar objects from unTimely

Utilizing a decade-long unTimely dataset, supplemented by multi-band data from archives, we search for young stellar objects (YSOs) with variations larger than one magnitude in W1 band within a region of 110 square degrees in the Galactic plane, covered by VISTA Variables in the Via Lactea (VVV). A total of 641 candidate YSOs have been identified. We classified them into bursts, dips, faders, seculars, and irregulars. Within the burst category, 18 sources were identified as FUor candidates and 1 as an EXor candidate. Irregulars are the most prevalent in the sample. In both bursts and faders, the redder sources tend to show a pattern of bluer when brighter, whereas the bluer sources display the opposite trend, possibly related to the accretion structure of YSOs at different stages. Finally, we obtained the recurrence time scale for FUor eruptions at various stages of YSO evolution. Our findings indicate that younger YSOs generally experience more frequent eruptions compared to older ones.

astro-ph.SR↗

arXiv2024

AT2023lli: A Tidal Disruption Event with Prominent Optical Early Bump and Delayed Episodic X-ray Emission

High-cadence, multiwavelength observations have continuously revealed the diversity of tidal disruption events (TDEs), thus greatly advancing our knowledge and understanding of TDEs. In this work, we conducted an intensive optical-UV and X-ray follow-up campaign of TDE AT2023lli, and found a remarkable month-long bump in its UV/optical light curve nearly two months prior to maximum brightness. The bump represents the longest separation time from the main peak among known TDEs to date. The main UV/optical outburst declines as $t^{-4.10}$, making it one of the fastest decaying optically selected TDEs. Furthermore, we detected sporadic X-ray emission 30 days after the UV/optical peak, accompanied by a reduction in the period of inactivity. It is proposed that the UV/optical bump could be caused by the self-intersection of the stream debris, whereas the primary peak is generated by the reprocessed emission of the accretion process. In addition, our results suggest that episodic X-ray radiation during the initial phase of decline may be due to the patched obscurer surrounding the accretion disk, a phenomenon associated with the inhomogeneous reprocessing process. The double TDE scenario, in which two stars are disrupted in sequence, is also a possible explanation for producing the observed early bump and main peak. We anticipate that the multicolor light curves of TDEs, especially in the very early stages, and the underlying physics can be better understood in the near future with the assistance of dedicated surveys such as the deep high-cadence survey of the 2.5-meter Wide Field Survey Telescope (WFST).

astro-ph.HE↗

arXiv2023

A Gigantic Mid-Infrared Outburst in an Embedded Class-I Young Stellar Object J064722.95+031644.6

We report the serendipitous discovery of a giant mid-infrared (MIR) outburst from a previously unknown source near a star-forming region in the constellation Monoceros. The source gradually brightened by a factor of 5 from 2014 to 2016 before an abrupt rise by a factor of more than 100 in 2017. A total amplitude increase of >500 at 4.5 microns has since faded by a factor of about 10. Prior to the outburst, it was only detected at wavelengths longer than 1.8 microns in UKIDSS, Spitzer, and Herschel with a spectral energy distribution of a Class I Young Stellar Object (YSO). It has not been detected in recent optical surveys, suggesting that it is deeply embedded. With a minimum distance of 3.5 kpc, the source has a bolometric luminosity of at least 9 $L_\odot$ in the quiescent state and 400 $L_\odot$ at the peak of the eruption. The maximum accretion rate is estimated to be at least a few $10^{-5}$ $M_\odot$ year$^{-1}$. It shares several common properties with another eruptive event, WISE~J142238.82-611553.7: exceptionally large amplitude, featureless near-infrared spectrum with the exception of H_2 lines, intermediate eruption duration, an embedded Class I YSO, and a low radiative temperature (<600-700 K) in outburst. We interpret that the radiation from the inner accretion disk and young star is obscured and reprocessed by either an inflated outer disk or thick dusty outflow on scales > 6.5 AU during the outburst.

astro-ph.SR↗

arXiv2023

Analysis of a stochastic SIR model with media effects

In this study, we investigate a stochastic SIR model with media effects. The uniqueness and the existence of a global positive solution are studied. The sufficient conditions of extinction and persistence of the disease are established. We obtain the basic reproduction number $R_0^S$ for stochastic system, which can act as the threshold given small environmental noise. Note that large noise can induce the disease extinction with probability of 1, suggesting that environmental noises can not be ignored when investigating threshold dynamics. Further, inclusion of media induced behaviour changes does not affect the threshold itself, which is similar to the conclusion of the deterministic models. However, numerical simulations suggest that media impacts induce the disease infection decline.

q-bio.PE↗

arXiv2023

Jac-PCG Based Low-Complexity Precoding for Extremely Large-Scale MIMO Systems

Extremely large-scale multiple-input-multipleoutput (XL-MIMO) has been reviewed as a promising technology for future sixth-generation (6G) networks to achieve higher performance. In practice, various linear precoding schemes, such as zero-forcing (ZF) and regularized ZF (RZF) precoding, are sufficient to achieve near-optimal performance in traditional massive MIMO (mMIMO) systems. It is critical to note that in large-scale antenna arrays the operation of channel matrix inversion poses a significant computational challenge for these precoders. Therefore, we explore several iterative methods for determining the precoding matrix for XL-MIMO systems instead of direct matrix inversion. Taking into account small- and large-scale fading as well as spatial correlation between antennas, we study their computational complexity and convergence rate. Furthermore, we propose the Jacobi-Preconditioning Conjugate Gradient (Jac-PCG) iterative inversion method, which enjoys a faster convergence speed than the CG method. Besides, the closed-form expression of spectral efficiency (SE) considering the interference between subarrays in downlink XL-MIMO systems is derived. In the numerical results, it is shown that the complexity given by the Jac-PCG algorithm has about 54% reduction than the traditional RZF algorithm at basically the same SE performance.

cs.IT↗

arXiv2018

Survey on Unmanned Aerial Vehicle Networks: A Cyber Physical System Perspective

Unmanned aerial vehicle (UAV) networks are playing an important role in various areas due to their agility and versatility, which have attracted significant attention from both the academia and industry in recent years. As an integration of the embedded systems with communication devices, computation capabilities and control modules, the UAV network could build a closed loop from data perceiving, information exchanging, decision making to the final execution, which tightly integrates the cyber processes into the physical devices. Therefore, the UAV network could be considered as a cyber physical system (CPS). Revealing the coupling effects among the three interacted components in this CPS system, i.e., communication, computation and control, is envisioned as the key to properly utilize all the available resources and hence improve the performance of the UAV networks. In this paper, we present a comprehensive survey on the UAV networks from a CPS perspective. Firstly, we respectively research the basics and advances with respect to the three CPS components in the UAV networks. Then we look inside to investigate how these components contribute to the system performance by classifying the UAV networks into three hierarchies, i.e., the cell level, the system level, and the system of system level. Further, the coupling effects among these CPS components are explicitly illustrated, which could be enlightening to deal with the challenges in each individual aspect. New research directions and open issues are discussed at the end of this survey. With this intensive literature review, we try to provide a novel insight into the state-of-the-art in the UAV networks.

cs.NI↗