Researcher profile

Zheng Wang

Zheng Wang contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author

96works

0followers

33topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

An Enigmatic PeVatron in an Area around HII Region G35.6$-$0.5

Identifying Galactic PeVatrons (PeV particle accelerators) from the ultra-high-energy (UHE, >100 TeV) $γ$-ray sources plays a crucial role in revealing the origin of Galactic cosmic rays. The UHE source 1LHAASO J1857+0203u is suggested to be associated with HESS J1858+020, which may be attributed to the possible PeVatron candidate supernova remnant (SNR) G35.6$-$0.4 or HII region G35.6$-$0.5. We perform detailed analysis on the very-high-energy and UHE $γ$-ray emissions towards this region with data from the Large High Altitude Air Shower Observatory (LHAASO). 1LHAASO J1857+0203u is detected with a significance of 11.6$σ$ above 100 TeV, indicating the presence of a PeVatron. It has an extension of $\sim 0.18^\circ$ with a power-law (PL) spectral index of $\sim$2.5 in 1-25 TeV and a point-like emission with a PL spectral index of $\sim$3.2 above 25 TeV. Using the archival CO and HI data, we identify some molecular and atomic clouds that may be associated with the TeV $γ$-ray emissions. Our modelling indicates that the TeV $γ$-ray emissions are unlikely to arise from the clouds illuminated by the protons that escaped from SNR G35.6$-$0.4. In the scenario that HII region G35.6$-$0.5 could accelerate particles to the UHE band, the observed GeV-TeV $γ$-ray emission could be well explained by a hadronic model with a PL spectral index of $\sim$2.0 and cutoff energy of $\sim$450 TeV. However, an evolved pulsar wind nebula origin cannot be ruled out.

preprint2026arXiv

An Ultrahigh-energy $γ$-ray Bubble Powered by a Super PeVatron

We report the detection of a $γ$-ray bubble spanning at least 100$\rm deg^2$ in ultra high energy (UHE) up to a few PeV in the direction of the star-forming region Cygnus X, implying the presence Super PeVatron(s) accelerating protons to at least 10 PeV. A log-parabola form with the photon index $Γ(E) = (2.71 \pm 0.02) + (0.11 \pm 0.02) \times \log_{10} (E/10 \ {\rm TeV})$ is found fitting the gamma-ray energy spectrum of the bubble well. UHE sources, `hot spots' correlated with very massive molecular clouds, and a quasi-spherical amorphous $γ$-ray emitter with a sharp central brightening are observed in the bubble. In the core of $\sim 0.5^{\circ}$, spatially associating with a region containing massive OB association (Cygnus OB2) and a microquasar (Cygnus X-3), as well as previously reported multi-TeV sources, an enhanced concentration of UHE $γ$-rays are observed with 2 photons at energies above 1 PeV. The general feature of the bubble, the morphology and the energy spectrum, are reasonably reproduced by the assumption of a particle accelerator in the core, continuously injecting protons into the ambient medium.

preprint2026arXiv

Beyond Prompts: Space-Time Decoupling Control-Plane Jailbreaks in LLM Structured Output

Content Warning: This paper may contain unsafe or harmful content generated by LLMs that may be offensive to readers. Large Language Models (LLMs) are extensively used as tooling platforms through structured output APIs to ensure syntax compliance so that robust integration with existing software, like agent systems, can be achieved. However, the feature enabling the functionality of grammar-guided structured output presents significant security vulnerabilities. In this work, we reveal a critical control-plane attack surface orthogonal to traditional data-plane vulnerabilities. We introduce Constrained Decoding Attack (CDA), a novel jailbreak class that weaponizes structured output constraints to bypass both external auditing and internal safety alignment. Unlike prior attacks focused on input prompt designs, CDA operates by embedding malicious intent in schema-level grammar rules (control-plane) while maintaining benign surface prompts (data-plane). We instantiate this with two proof-of-concept attacks: EnumAttack, which embeds malicious content in enum fields; and the more evasive DictAttack, which decouples the malicious payload across a benign prompt and a dictionary-based grammar. Our evaluation spans a broad spectrum of 13 proprietary/open-weight models. In particular, DictAttack achieves 94.3--99.5% ASR across five benchmarks on gpt-5, gemini-2.5-pro, deepseek-r1, and gpt-oss-120b. Furthermore, we demonstrate the significant challenge in defending against these threats: while basic grammar auditing mitigates EnumAttack, the more sophisticated DictAttack maintains a 75.8% ASR even against multiple state-of-the-art jailbreak guardrails. This exposes a critical "semantic gap" in current safety architectures and underscores the urgent need for cross-plane defenses that can bridge the data and control planes to secure the LLM generation pipeline.

preprint2026arXiv

Beyond Seen Bounds: Class-Centric Polarization for Single-Domain Generalized Deep Metric Learning

Single-domain generalized deep metric learning (SDG-DML) faces the dual challenge of both category and domain shifts during testing, limiting real-world applications. Therefore, aiming to learn better generalization ability on both unseen categories and domains is a realistic goal for the SDG-DML task. To deliver the aspiration, existing SDG-DML methods employ the domain expansion-equalization strategy to expand the source data and generate out-of-distribution samples. However, these methods rely on proxy-based expansion, which tends to generate samples clustered near class proxies, failing to simulate the broad and distant domain shifts encountered in practice. To alleviate the problem, we propose CenterPolar, a novel SDG-DML framework that dynamically expands and constrains domain distributions to learn a generalizable DML model for wider target domain distributions. Specifically, \textbf{CenterPolar} contains two collaborative class-centric polarization phases: (1) Class-Centric Centrifugal Expansion ($C^3E$) and (2) Class-Centric Centripetal Constraint ($C^4$). In the first phase, $C^3E$ drives the source domain distribution by shifting the source data away from class centroids using centrifugal expansion to generalize to more unseen domains. In the second phase, to consolidate domain-invariant class information for the generalization ability to unseen categories, $C^4$ pulls all seen and unseen samples toward their class centroids while enforcing inter-class separation via centripetal constraint. Extensive experimental results on widely used CUB-200-2011 Ext., Cars196 Ext., DomainNet, PACS, and Office-Home datasets demonstrate the superiority and effectiveness of our CenterPolar over existing state-of-the-art methods. The code will be released after acceptance.

preprint2026arXiv

Constraining the Cosmic-ray Energy Based on Observations of Nearby Galaxy Clusters by LHAASO

Galaxy clusters act as reservoirs of high-energy cosmic rays (CRs). As CRs propagate through the intracluster medium, they generate diffuse $γ$-rays detectable by arrays such as LHAASO. These $γ$-rays result from proton-proton ($pp$) collisions of very high-energy cosmic rays (VHECRs) or inverse Compton (IC) scattering of positron-electron pairs created by $pγ$ interactions of ultra-high-energy cosmic rays (UHECRs). We analyzed diffuse $γ$-ray emission from the Coma, Perseus, and Virgo clusters using LHAASO data. Diffuse emission was modeled as a disk of radius $R_{500}$ for each cluster while accounting for point sources. No significant diffuse emission was detected, yielding 95\% confidence level (C.L.) upper limits on the $γ$-ray flux: for WCDA (1-25~TeV) and KM2A ($>25$~TeV), less than $(49.4, 13.7, 54.0)$ and $(1.34, 1.14, 0.40) \times 10^{-14}$~ph~cm$^{-2}$~s$^{-1}$ for Coma, Perseus, and Virgo, respectively. The $γ$-ray upper limits can be used to derive model-independent constraints on the integral energy of CRp above 10~TeV (corresponding to the LHAASO observational range $>1$~TeV under the $pp$ scenario) to be less than $(1.96, 0.59, 0.08) \times 10^{61}$~erg. The absence of detectable annuli/ring-like structures, indicative of cluster accretion or merging shocks, imposes further constraints on models in which the UHECRs are accelerated in the merging shocks of galaxy clusters.

preprint2026arXiv

Constraints on heavy decaying dark matter from 570 days of LHAASO observations

The Kilometer Square Array~(KM2A) of the Large High Altitude Air Shower Observatory (LHAASO) aims at surveying the northern gamma-ray sky at energies above 10 TeV with unprecedented sensitivity. Gamma-ray observations have long been one of the most powerful tools for dark matter searches, as e.g., high-energy gamma-rays could be produced by the decays of heavy dark matter particles. In this letter, we present the first dark matter analysis with LHAASO-KM2A, using the first 340~days of data from 1/2-KM2A and 230~days of data from 3/4-KM2A. Several regions of interest are used to search for a signal and account for the residual cosmic-ray background after gamma/hadron separation. We find no excess of dark matter signals, and thus place some of the strongest gamma-ray constraints on the lifetime of heavy dark matter particles with mass between 10^5 and 10^9~GeV. Our results with LHAASO are robust, and have important implications for dark matter interpretations of the diffuse astrophysical high-energy neutrino emission.

preprint2026arXiv

Deep view of Composite SNR CTA1 with LHAASO in $γ$-rays up to 300 TeV

The ultra-high-energy (UHE) gamma-ray source 1LHAASO J0007+7303u is positionally associated with the composite SNR CTA1 that is located at high Galactic Latitude $b\approx 10.5^\circ$. This provides a rare opportunity to spatially resolve the component of the pulsar wind nebula (PWN) and supernova remnant (SNR) at UHE. This paper conducted a dedicated data analysis of 1LHAASO J0007+7303u using the data collected from December 2019 to July 2023. This source is well detected with significances of 21$σ$ and 17$σ$ at 8$-$100 TeV and $>$100 TeV, respectively. The corresponding extensions are determined to be 0.23$^{\circ}\pm$0.03$^{\circ}$ and 0.17$^{\circ}\pm$0.03$^{\circ}$. The emission is proposed to originate from the relativistic electrons and positrons accelerated within the PWN of PSR J0007+7303. The energy spectrum is well described by a power-law with an exponential cutoff function $dN/dE = (42.4\pm4.1)(\frac{E}{20\rm\ TeV})^{-2.31\pm0.11}\exp(-\frac{E}{110\pm25\rm\ TeV})$ $\rm\ TeV^{-1}\ cm^{-2}\ s^{-1}$in the energy range from 8 TeV to 300 TeV, implying a steady-state parent electron spectrum $dN_e/dE_e\propto (\frac{E_e}{100\rm\ TeV})^{-3.13\pm0.16}\exp[(\frac{-E_e}{373\pm70\rm\ TeV})^2]$ at energies above $\approx 50 \rm\ TeV$. The cutoff energy of the electron spectrum is roughly equal to the expected current maximum energy of particles accelerated at the PWN terminal shock. Combining the X-ray and gamma-ray emission, the current space-averaged magnetic field can be limited to $\approx 4.5\rm\ μG$. To satisfy the multi-wavelength spectrum and the $γ$-ray extensions, the transport of relativistic particles within the PWN is likely dominated by the advection process under the free-expansion phase assumption.

preprint2026arXiv

Discovery of a new $γ$-ray source LHAASO J0341+5258 with emission up to 200TeV

We report the discovery of a new unidentified extended $γ$-ray source in the Galactic plane named LHAASO J0341+5258 with a pre-trial significance of 8.2 standard deviations above 25 TeV. The best fit position is R.A.$=55.34^{\circ}\pm0.11^{\circ}$ and Dec$=52.97^{\circ}\pm0.07^{\circ}$. The angular size of LHAASO J0341+5258 is $0.29^\circ \pm 0.06^\circ_{stat} \pm0.02^\circ_{sys}$. The flux above 25 TeV is about $20\%$ of the flux of Crab Nebula. Although a power-law fit of the spectrum from 10 TeV to 200 TeV with the photon index $α=2.98 \pm 0.19_{stat} \pm 0.02_{sys}$ is not excluded, the LHAASO data together with the flux upper limit at 10 GeV set by the Fermi LAT observation, indicate a noticeable steepening of an initially hard power-law spectrum %($α\leq 1.75$) spectrum with a cutoff at $\approx 50$ TeV. We briefly discuss the origin of UHE gamma-rays. The lack of an energetic pulsar and a young SNR inside or in the vicinity of LHAASO J0341+5258 challenge, but do not exclude both the leptonic and hadronic scenarios of gamma-ray production.

preprint2026arXiv

Discovery of the Ultra-high energy gamma-ray source LHAASO J2108+5157

We report the discovery of a UHE gamma-ray source, LHAASO J2108+5157, by analyzing the LHAASO-KM2A data of 308.33 live days. Significant excess of gamma-ray induced showers is observed in both energy bands of 25-100 TeV and $\gt$100 TeV with 9.5 sigma and 8.5 sigma, respectively. This source is not significantly favored as an extensive source with the angular extension smaller than the point-spread function of KM2A. The measured energy spectrum from 20 to 200 TeV can be approximately described by a power-law function with an index of -2.83$\pm$ 0.18stat. A harder spectrum is demanded at lower energies considering the flux upper limit set by Fermi-LAT observations. The position of the gamma-ray emission is correlated with a giant molecular cloud, which favors a hadronic origin. No obvious counterparts have been found, deeper multiwavelength observations will help to shed new light on this intriguing UHE source.

preprint2026arXiv

Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning

In this report, we introduce DASD-4B-Thinking, a lightweight yet highly capable, fully open-source reasoning model. It achieves SOTA performance among open-source models of comparable scale across challenging benchmarks in mathematics, scientific reasoning, and code generation -- even outperforming several larger models. We begin by critically reexamining a widely adopted distillation paradigm in the community: SFT on teacher-generated responses, also known as sequence-level distillation. Although a series of recent works following this scheme have demonstrated remarkable efficiency and strong empirical performance, they are primarily grounded in the SFT perspective. Consequently, these approaches focus predominantly on designing heuristic rules for SFT data filtering, while largely overlooking the core principle of distillation itself -- enabling the student model to learn the teacher's full output distribution so as to inherit its generalization capability. Specifically, we identify three critical limitations in current practice: i) Inadequate representation of the teacher's sequence-level distribution; ii) Misalignment between the teacher's output distribution and the student's learning capacity; and iii) Exposure bias arising from teacher-forced training versus autoregressive inference. In summary, these shortcomings reflect a systemic absence of explicit teacher-student interaction throughout the distillation process, leaving the essence of distillation underexploited. To address these issues, we propose several methodological innovations that collectively form an enhanced sequence-level distillation training pipeline. Remarkably, DASD-4B-Thinking obtains competitive results using only 448K training samples -- an order of magnitude fewer than those employed by most existing open-source efforts. To support community research, we publicly release our models and the training dataset.

preprint2026arXiv

Energy calibration of LHAASO-KM2A using the cosmic ray Moon shadow

We present a precise measurement of the westward, rigidity-dependent shift of the Moon's shadow using three and a half years of cosmic-ray data collected by the Kilometer Square Array (KM2A) of the Large High Altitude Air Shower Observatory (LHAASO). These measurements enable us to calibrate the detector energy response in the range 20-260 TeV, with results showing excellent agreement with the response derived from Monte Carlo (MC) simulations of the KM2A detector. We also measure a best-fit parameter $ε= 0.015 \pm 0.08$, corresponding to a 95% confidence interval of [-14%, +17%] for the energy-scale estimation. This result establishes the exceptional accuracy of the KM2A-MC in simulating the detector's response within this energy range.

preprint2026arXiv

Energy-Dependent Shifts of Medium-Scale Anisotropies in Very-High-Energy Cosmic Rays Observed by LHAASO-KM2A

Small deviations from isotropy in the arrival directions of Galactic cosmic rays serve as a unique probe of the local magnetic environment. In this Letter, we report observations of medium-scale anisotropies (MSA) at energies above 10 TeV using the LHAASO-KM2A array. Our analysis identifies four regions of excess and four regions of deficit, each spanning angular scales of approximately ten degrees. Crucially, we detect significant energy-dependent shifts in the centroids of two excess regions: Region B and the newly identified Region $\mathrm{\widetilde{D}}$. We also characterize the energy evolution of the fractional relative intensity across both excess and deficit regions. These findings imply that the observed anisotropies are shaped by the specific realization of the local turbulent magnetic field within the cosmic ray scattering length. Such energy-dependent behaviors impose strict constraints on local turbulence models and cosmic ray propagation theories.

preprint2026arXiv

Evidence for particle acceleration approaching PeV energies in the W51 complex

The $γ$-ray emission from the W51 complex is widely acknowledged to be attributed to the interaction between the cosmic rays (CRs) accelerated by the shock of supernova remnant (SNR) W51C and the dense molecular clouds in the adjacent star-forming region, W51B. However, the maximum acceleration capability of W51C for CRs remains elusive. Based on observations conducted with the Large High Altitude Air Shower Observatory (LHAASO), we report a significant detection of $γ$ rays emanating from the W51 complex, with energies from 2 TeV to 200 TeV. The LHAASO measurements, for the first time, extend the $γ$-ray emission from the W51 complex beyond 100 TeV and reveal a significant spectrum bending at tens of TeV. By combining the ``$π^0$-decay bump" featured data from Fermi-LAT, the broadband $γ$-ray spectrum of the W51 region can be well-characterized by a simple pp-collision model. The observed spectral bending feature suggests an exponential cutoff at $\sim400$~TeV or a power-law break at $\sim200$~TeV in the CR proton spectrum, most likely providing the first evidence of SNRs serving as CR accelerators approaching the PeV regime. Additionally, two young star clusters within W51B could also be theoretically viable to produce the most energetic $γ$ rays observed by LHAASO. Our findings strongly support the presence of extreme CR accelerators within the W51 complex and provide new insights into the origin of Galactic CRs.

preprint2026arXiv

Exploring Lorentz Invariance Violation from Ultra-high-energy Gamma Rays Observed by LHAASO

Recently the LHAASO Collaboration published the detection of 12 ultra-high-energy gamma-ray sources above 100 TeV, with the highest energy photon reaching 1.4 PeV. The first detection of PeV gamma rays from astrophysical sources may provide a very sensitive probe of the effect of the Lorentz invariance violation (LIV), which results in decay of high-energy gamma rays in the superluminal scenario and hence a sharp cutoff of the energy spectrum. Two highest energy sources are studied in this work. No signature of the existence of LIV is found in their energy spectra, and the lower limits on the LIV energy scale are derived. Our results show that the first-order LIV energy scale should be higher than about 10^5 times the Planck scale M_{pl} and that the second-order LIV scale is >10^{-3}M_{pl}. Both limits improve by at least one order of magnitude the previous results.

preprint2026arXiv

Extended Very-High-Energy Gamma-Ray Emission Surrounding PSR J0622 + 3749 Observed by LHAASO-KM2A

We report the discovery of an extended very-high-energy (VHE) gamma-ray source around the location of the middle-aged (207.8 kyr) pulsar PSR J0622+3749 with the Large High Altitude Air Shower Observatory (LHAASO). The source is detected with a significance of $8.2σ$ for $E>25$~TeV assuming a Gaussian template. The best-fit location is (R.A., Dec.)$=(95^{\circ}\!.47\pm0^{\circ}\!.11,\,37^{\circ}\!.92 \pm0^{\circ}\!.09)$, and the extension is $0^{\circ}\!.40\pm0^{\circ}\!.07$. The energy spectrum can be described by a power-law spectrum with an index of ${-2.92 \pm 0.17_{\rm stat} \pm 0.02_{\rm sys} }$. No clear extended multi-wavelength counterpart of the LHAASO source has been found from the radio to sub-TeV bands. The LHAASO observations are consistent with the scenario that VHE electrons escaped from the pulsar, diffused in the interstellar medium, and scattered the interstellar radiation field. If interpreted as the pulsar halo scenario, the diffusion coefficient, inferred for electrons with median energies of $\sim160$~TeV, is consistent with those obtained from the extended halos around Geminga and Monogem and much smaller than that derived from cosmic ray secondaries. The LHAASO discovery of this source thus likely enriches the class of so-called pulsar halos and confirms that high-energy particles generally diffuse very slowly in the disturbed medium around pulsars.

preprint2026arXiv

First Submillimeter Lights from Dome A: Tracing the Carbon Cycle in the Feedback of Massive Stars

The cycling of carbon between its ionized, atomic, and molecular phases shapes the chemical compositions and physical conditions of the interstellar medium (ISM). However, ground-based studies of the full carbon cycle have been limited by atmospheric absorption. Dome~A, the most promising site for submillimeter astronomy, has long resisted successful submillimeter astronomical observations. Using the 60~cm Antarctic Terahertz Explorer, we present the first successful CO ($4-3$) and [CI] ($^3P_1 - ^3P_0$) mapping observations of two archetypal triggered massive star-formation regions at Dome~A. These data, together with archival [CII], provide the first complete characterization of all three carbon phases in these environments. We find elevated C$^{0}$/CO abundance ratios in high-extinction regions, plausibly driven by deep penetration of intense radiation fields from massive stars into a clumpy ISM. These findings mark a major milestone for submillimeter astronomy at Dome~A and offer valuable insights into the impact of massive star feedback on the surrounding ISM.

preprint2026arXiv

From Context to Skills: Can Language Models Learn from Context Skillfully?

Many real-world tasks require language models (LMs) to reason over complex contexts that exceed their parametric knowledge. This calls for context learning, where LMs directly learn relevant knowledge from the given context. An intuitive solution is inference-time skill augmentation: extracting the rules and procedures from context into natural-language skills. However, constructing such skills for context learning scenarios faces two challenges: the prohibitive cost of manual skill annotation for long, technically dense contexts, and the lack of external feedback for automated skill construction. In this paper, we propose Ctx2Skill, a self-evolving framework that autonomously discovers, refines, and selects context-specific skills without human supervision or external feedback. At its core, a multi-agent self-play loop has a Challenger that generates probing tasks and rubrics, a Reasoner that attempts to solve them guided by an evolving skill set, and a neutral Judge that provides binary feedback. Crucially, both the Challenger and the Reasoner evolve through accumulated skills: dedicated Proposer and Generator agents analyze failure cases and synthesize them into targeted skill updates for both sides, enabling automated skill discovery and refinement. To prevent adversarial collapse caused by increasingly extreme task generation and over-specialized skill accumulation, we further introduce a Cross-time Replay mechanism that identifies the skill set achieving the best balance across representative cases for the Reasoner side, ensuring robust and generalizable skill evolution. The resulting skills can be plugged into any language model to obtain better context learning capability. Evaluated on four context learning tasks from CL-bench, Ctx2Skill consistently improves solving rates across backbone models.

preprint2026arXiv

LHAASO Detection of Ultra-High-Energy Gamma-Ray Emission toward the Giant Molecular Clouds

The $γ$-ray from Giant molecular clouds (GMCs) is regarded as the most ideal tool to perform in-situ measurement of cosmic ray (CR) density and spectra in our Galaxy. We report the first detection of $γ$-ray emissions in the very-high-energy (VHE) domain from the five nearby GMCs with a stacking analysis based on a 4.5-year $γ$-ray observation with the Large High Altitude Air Shower Observatory (LHAASO) experiment. The spectral energy distributions derived from the GMCs are consistent with the expected $γ$-ray flux produced via CR interacting with the ISM in the energy interval 1 - 100 $~\rm$ TeV. In addition, we investigate the presence of the CR spectral `knee' by introducing a spectral break in the $γ$-ray data. While no significant evidence for the CR knee is found, the current KM2A measurements from GMCs strongly favor a proton CR knee located above 0.9$~\rm$ PeV, which is consistent with the latest measurement of the CR spectrum by ground-based experiments.

preprint2026arXiv

Measurement of attenuation length of the muon content in extensive air showers from 0.3 to 30 PeV with LHAASO

The attenuation length of the muon content in extensive air showers provides important information regarding the generation and development of air showers. This information can be used not only to improve the description of such showers but also to test fundamental models of hadronic interactions. Using data from the LHAASO-KM2A experiment, the development of the muon content in high-energy air showers was studied. The attenuation length of muon content in the air showers was measured from experimental data in the energy range from 0.3 to 30 PeV using the constant intensity cut method. By comparing the attenuation length of the muon content with predictions from high-energy hadronic interaction models (QGSJET-II-04, SIBYLL 2.3d, and EPOS-LHC), it is evident that LHAASO results are significantly shorter than those predicted by the first two models (QGSJET-II-04 and SIBYLL 2.3d) but relatively close to those predicted by the third model (EPOS-LHC). Thus, the LHAASO data favor the EPOS-LHC model over the other two models. The three interaction models confirmed an increasing trend in the attenuation length as the cosmic-ray energy increases.

preprint2026arXiv

Measurement of Very-high-energy Diffuse Gamma-ray Emissions from the Galactic Plane with LHAASO-WCDA

The diffuse Galactic gamma-ray emission is a very important tool used to study the propagation and interaction of cosmic rays in the Milky Way. In this work, we report the measurements of the diffuse emission from the Galactic plane, covering Galactic longitudes from $15^{\circ}$ to $235^{\circ}$ and latitudes from $-5^{\circ}$ to $+5^{\circ}$, in an energy range of 1 TeV to 25 TeV, with the Water Cherenkov Detector Array (WCDA) of the Large High Altitude Air Shower Observatory (LHAASO). After masking the sky regions of known sources, the diffuse emission is detected with $24.6σ$ and $9.1σ$ significance in the inner Galactic plane and outer Galactic plane, respectively. The WCDA spectra in both regions can be well described by a power-law function, with spectral indices of $-2.67\pm0.05_{\rm stat}$ in the inner region and $-2.83\pm0.19_{\rm stat}$ in the outer region, respectively. Combined with the Square Kilometer Array (KM2A) measurements at higher energies, a clear softening of the spectrum is found in the inner region, with change of spectral indices by $\sim0.5$ at a break energy around $30$ TeV. The fluxes of the diffuse emission are higher by a factor of $1.5-2.7$ than the model prediction assuming local CR spectra and the gas column density, which are consistent with those measured by the KM2A. Along Galactic longitude, the spatial distribution of the diffuse emission shows deviation from that of the gas column density. The spectral shape of the diffuse emission are possibly variation in different longitude region. The WCDA measurements bridge the gap between the low-energy measurements by space detectors and the ultra-high-energy observations by LHAASO-KM2A and other experiments. These results suggest that improved modeling of the wide-band diffuse emission is required.

preprint2026arXiv

Modality-Aware Identity Construction and Counterfactual Structure Learning for ID-Free Multimodal Recommendation

Multimodal recommendation has attracted extensive attention by leveraging heterogeneous modality information to alleviate data sparsity and improve recommendation accuracy. Existing methods have attempted to replace ID embeddings with multimodal features and have achieved promising preliminary results. However, these methods still exhibit the following two limitations: (1) the reconstructed ID representations remain relatively static and fail to fully exploit multimodal semantics; and (2) the graph learning process is insufficient in mining latent long-tail semantic relations and is easily affected by popularity bias. To address these issues, we propose a novel method named Modality-Aware Identity Construction and Counterfactual Structure Learning for ID-free Multimodal Recommendation (MAIL). Specifically, we design a modality-aware identity construction module that dynamically modulates positional encodings with multimodal semantics to construct content-aware ID-free identity representations. Then, we propose a counterfactual structure learning paradigm that mines low-exposure semantic neighbors via popularity penalization and alleviates popularity bias. Extensive experiments are conducted on five public Amazon datasets. Experimental results show that MAIL achieves average improvements of 7.81% in Recall@10 and 12.81% in NDCG@10 compared with the baseline models. Our code is available at https://github.com/HubuKG/MAIL.

preprint2026arXiv

Reflections and New Directions for Human-Centered Large Language Models

Large Language Models (LLMs) are increasingly shaping the private and professional lives of users, with numerous applications in business, education, finance, healthcare, law, and science. With this rise in global influence comes greater urgency to build, evaluate, and deploy these systems in a manner that prioritizes not only technical capabilities but also human priorities. This work presents a framework for developing Human-Centered Large Language Models (HCLLMs), which integrates perspectives from Natural Language Processing (NLP), Human-Computer Interaction (HCI), and responsible AI. Considering the ethics, economics, and technical objectives of language modeling, we argue that model developers need to address human concerns, preferences, values, and goals, not only during a cursory post-training stage, but rather with rigor and care at every stage of the pipeline. This paper offers human-centered insights and recommendations for developers at each stage, from system design to data sourcing, model training, evaluation, and responsible deployment. Then we conclude with a case study, applying these insights to understand the future of work with HCLLMs.

preprint2026arXiv

Reinforced Collaboration in Multi-Agent Flow Networks

Multi-agent systems provide a powerful way to extend large language models (LLMs) by decomposing a complex task into specialized subtasks handled by different agents. However, their performance is often hindered by error propagation, arising from suboptimal workflow design or inaccurate agent outputs, which can propagate through the agent collaboration process and degrade final results. To address the challenges, we present MANGO (Multi-Agent Network Gradient Optimization), a data-driven framework that organizes and refines agent collaboration via a flow network constructed from past successful workflows. MANGO integrates reinforcement learning and textual gradients to jointly optimize workflow paths and agent behaviors, while a skipping mechanism prevents redundant updates to well-optimized agents for improving efficiency. Extensive experiments on seven benchmarks show that MANGO achieves up to 12.8% performance improvement over state-of-the-art baselines, enhances efficiency by 47.4%, and generalizes effectively to unseen domains. Our code and datasets are publicly available at https://github.com/openJiuwen-ai/agent-store/tree/main/community/mango.

preprint2026arXiv

SuperEar: Eavesdropping on Mobile Voice Calls via Stealthy Acoustic Metamaterials

Acoustic eavesdropping is a privacy risk, but existing attacks rarely work in real outdoor situations where people make phone calls on the move. We present SuperEar, the first portable system that uses acoustic metamaterials to reliably capture conversations in these scenarios. We show that the threat is real as a practical prototype can be implemented to enhance faint signals, cover the full range of speech with a compact design, and reduce noise and distortion to produce clear audio. We show that SuperEar can be implemented from low-cost 3D-printed parts and off-the-shelf hardware. Experimental results show that SuperEar can recover phone call audio with a success rate of over 80% at distances of up to 4.6 m - more than twice the range of previous approaches. Our findings highlight a new class of privacy threats enabled by metamaterial technology that requires attention.

preprint2026arXiv

The New Compiler Stack: A Survey on the Synergy of LLMs and Compilers

This survey has provided a systematic overview of the emerging field of LLM-enabled compilation by addressing several key research questions. We first answered how LLMs are being integrated by proposing a comprehensive, multi-dimensional taxonomy that categorizes works based on their Design Philosophy (Selector, Translator, Generator), LLM Methodology, their operational Level of Code Abstraction, and the specific Task Type they address. In answering what advancements these approaches offer, we identified three primary benefits: the democratization of compiler development, the discovery of novel optimization strategies, and the broadening of the compiler's traditional scope. Finally, in addressing the field's challenges and opportunities, we highlighted the critical hurdles of ensuring correctness and achieving scalability, while identifying the development of hybrid systems as the most promising path forward. By providing these answers, this survey serves as a foundational roadmap for researchers and practitioners, charting the course for a new generation of LLM-powered, intelligent, adaptive and synergistic compilation tools.

preprint2026arXiv

Transient Large-Scale Anisotropy in TeV Cosmic Rays due to an Interplanetary Coronal Mass Ejection

Large- or medium-scale cosmic ray anisotropy at TeV energies has not previously been confirmed to vary with time. Transient anisotropy changes have been observed below 150 GeV, especially near the passage of an interplanetary shock and coronal mass ejection containing a magnetic flux rope ejected by a solar storm, which can trigger a geomagnetic storm with practical consequences. In such events, cosmic rays provide remote sensing of the magnetic field properties. Here we report the observation of transient large-scale anisotropy in TeV cosmic ray ions using data from the Large High Altitude Air Shower Observatory (LHAASO). We analyze hourly skymaps of the transient cosmic ray intensity excess or deficit, the gradient of which indicates the direction and magnitude of transient large-scale anisotropy across the field of view. We observe enhanced anisotropy above typical hourly fluctuations with $>$5$σ$ significance during some hours of November 4, 2021, in separate data sets for four primary cosmic ray energy ranges of median energy from $E$=0.7 to 3.1 TeV. The gradient varies with energy as $E^γ$, where $γ\approx-0.5$. At a median energy $\leq$1.0 TeV, this gradient corresponds to a dipole anisotropy of at least 1\%, or possibly a weaker anisotropy of higher order. This new type of observation opens the opportunity to study interplanetary magnetic structures using air shower arrays around the world, complementing existing in situ and remote measurements of plasma properties.

preprint2025arXiv

CREPES-X: Hierarchical Bearing-Distance-Inertial Direct Cooperative Relative Pose Estimation System

Relative localization is critical for cooperation in autonomous multi-robot systems. Existing approaches either rely on shared environmental features or inertial assumptions or suffer from non-line-of-sight degradation and outliers in complex environments. Robust and efficient fusion of inter-robot measurements such as bearings, distances, and inertials for tens of robots remains challenging. We present CREPES-X (Cooperative RElative Pose Estimation System with multiple eXtended features), a hierarchical relative localization framework that enhances speed, accuracy, and robustness under challenging conditions, without requiring any global information. CREPES-X starts with a compact hardware design: InfraRed (IR) LEDs, an IR camera, an ultra-wideband module, and an IMU housed in a cube no larger than 6cm on each side. Then CREPES-X implements a two-stage hierarchical estimator to meet different requirements, considering speed, accuracy, and robustness. First, we propose a single-frame relative estimator that provides instant relative poses for multi-robot setups through a closed-form solution and robust bearing outlier rejection. Then a multi-frame relative estimator is designed to offer accurate and robust relative states by exploring IMU pre-integration via robocentric relative kinematics with loosely- and tightly-coupled optimization. Extensive simulations and real-world experiments validate the effectiveness of CREPES-X, showing robustness to up to 90% bearing outliers, proving resilience in challenging conditions, and achieving RMSE of 0.073m and 1.817° in real-world datasets.

preprint2025arXiv

Observation of the $γ$-ray Emission from W43 with LHAASO

In this paper, we report the detection of the very-high-energy (VHE, $ 100{\rm\ GeV} < E < 100{\rm\ TeV} $) and ultra-high-energy (UHE, $E > 100\rm\ TeV$) $γ$-ray emissions from the direction of the young star-forming region W43, observed by the Large High Altitude Air Shower Observation (LHAASO). The extended $γ$-ray source was detected with a significance of ${\sim}16\,σ$ by KM2A and ${\sim}17\,σ$ by WCDA, respectively. The angular extension of this $γ$-ray source is about 0.5 degrees, corresponding to a physical size of about 50 pc. We discuss the origin of the $γ$-ray emission and possible cosmic ray acceleration in the W43 region using multi-wavelength data. Our findings suggest that W43 is likely another young star cluster capable of accelerating cosmic rays (CRs) to at least several hundred TeV.

preprint2025arXiv

RAJ-PGA: Reasoning-Activated Jailbreak and Principle-Guided Alignment Framework for Large Reasoning Models

Large Reasoning Models (LRMs) face a distinct safety vulnerability: their internal reasoning chains may generate harmful content even when the final output appears benign. To address this overlooked risk, we first propose a novel attack paradigm, Reasoning-Activated Jailbreak (RAJ) via Concretization, which demonstrates that refining malicious prompts to be more specific can trigger step-by-step logical reasoning that overrides the model's safety protocols. To systematically mitigate this vulnerability, we further develop a scalable framework for constructing high-quality safety alignment datasets. This framework first leverages the RAJ attack to elicit challenging harmful reasoning chains from LRMs, then transforms these high-risk traces into safe, constructive, and educational responses through a tailored Principle-Guided Alignment (PGA) mechanism. Then, we introduce the PGA dataset, a verified alignment dataset containing 3,989 samples using our proposed method. Extensive experiments show that fine-tuning LRMs with PGA dataset significantly enhances model safety, achieving up to a 29.5% improvement in defense success rates across multiple jailbreak benchmarks. Critically, our approach not only defends against sophisticated reasoning-based attacks but also preserves, even enhances, the model's general reasoning capabilities. This work provides a scalable and effective pathway for safety alignment in reasoning-intensive AI systems, addressing the core trade-off between safety and functional performance.

preprint2025arXiv

Study of Ultra-High-Energy Gamma-Ray Source 1LHAASO J0056+6346u and Its Possible Origins

We report a dedicated study of the newly discovered extended UHE $γ$-ray source 1LHAASO J0056+6346u. Analyzing 979 days of LHAASO-WCDA data and 1389 days of LHAASO-KM2A data, we observed a significant excess of $γ$-ray events with both WCDA and KM2A. Assuming a point power-law source with a fixed spectral index, the significance maps reveal excesses of ${\sim}12.65\,σ$, ${\sim}22.18\,σ$, and ${\sim}10.24\,σ$ in the energy ranges of 1--25 TeV, 25--100 TeV, and $> 100$ TeV, respectively. We use a 3D likelihood algorithm to derive the morphological and spectral parameters, and the source is detected with significances of $12.65\,σ$ by WCDA and $25.27\,σ$ by KM2A. The best-fit positions derived from WCDA and KM2A data are (R.A. = $13.96^\circ\pm0.09^\circ$, Decl. = $63.92^\circ\pm0.05^\circ$) and (R.A. = $14.00^\circ\pm0.05^\circ$, Decl. = $63.79^\circ\pm0.02^\circ$), respectively. The angular size ($r_{39}$) of 1LHAASO J0056+6346u is $0.34^\circ\pm0.04^\circ$ at 1--25 TeV and $0.24^\circ\pm0.02^\circ$ at $> 25$ TeV. The differential flux of this UHE $γ$-ray source can be described by an exponential cutoff power-law function: $(2.67\pm0.25) \times 10^{-15} (E/20\,\text{TeV})^{-1.97\pm0.10} e^{-E/(55.1\pm7.2)\,\text{TeV}} \,\text{TeV}^{-1}\,\text{cm}^{-2}\,\text{s}^{-1}$. To explore potential sources of $γ$-ray emission, we investigated the gas distribution around 1LHAASO J0056+6346u. 1LHAASO J0056+6346u is likely to be a TeV PWN powered by an unknown pulsar, which would naturally explain both its spatial and spectral properties. Another explanation is that this UHE $γ$-ray source might be associated with gas content illuminated by a nearby CR accelerator, possibly the SNR candidate G124.0+1.4.

preprint2025arXiv

Ultrahigh-Energy Gamma-ray Emission Associated with Black Hole-Jet Systems

Black holes (BH), one of the most intriguing objects in the universe, can manifest themselves through electromagnetic radiation initiated by the accretion flow. Some stellar-mass BHs drive relativistic jets when accreting matter from their companion stars, forming microquasars. Non-thermal emission from the radio to tera-electronvolt (TeV) gamma-ray band has been observed from microquasars, indicating the acceleration of relativistic particles. Here we report detection of four microquasars (SS 433, V4641 Sgr, GRS 1915+105, MAXI J1820+070) of spectrum extending to the ultrahigh-energy (UHE; photon energy $E>100$ TeV) band and one microquasar (Cygnus X-1) of spectrum approaching 100 TeV, using the Large High Altitude Air Shower Observatory (LHAASO). Notably, the total emission associated with SS 433 cannot be interpreted with a single leptonic component. In the UHE band, its emission is in spatial coincidence with a giant atomic cloud, which is consistent with a hadronic origin. An elongated source is discovered from V4641 Sgr with the spectrum continuing up to 800 TeV. The detection of UHE gamma rays demonstrates that accreting BHs and their environments can operate as extremely efficient accelerators of particles out of 1 peta-electronvolt (PeV), suggesting microquasars to be important contributors to Galactic cosmic rays especially around the `knee' region.

preprint2025arXiv

Yggdrasil: Bridging Dynamic Speculation and Static Runtime for Latency-Optimal Tree-Based LLM Decoding

Speculative decoding improves LLM inference by generating and verifying multiple tokens in parallel, but existing systems suffer from suboptimal performance due to a mismatch between dynamic speculation and static runtime assumptions. We present Yggdrasil, a co-designed system that enables latency-optimal speculative decoding through context-aware tree drafting and compiler-friendly execution. Yggdrasil introduces an equal-growth tree structure for static graph compatibility, a latency-aware optimization objective for draft selection, and stage-based scheduling to reduce overhead. Yggdrasil supports unmodified LLMs and achieves up to $3.98\times$ speedup over state-of-the-art baselines across multiple hardware setups.

preprint2022arXiv

A Survey on Legal Judgment Prediction: Datasets, Metrics, Models and Challenges

Legal judgment prediction (LJP) applies Natural Language Processing (NLP) techniques to predict judgment results based on fact descriptions automatically. Recently, large-scale public datasets and advances in NLP research have led to increasing interest in LJP. Despite a clear gap between machine and human performance, impressive results have been achieved in various benchmark datasets. In this paper, to address the current lack of comprehensive survey of existing LJP tasks, datasets, models and evaluations, (1) we analyze 31 LJP datasets in 6 languages, present their construction process and define a classification method of LJP with 3 different attributes; (2) we summarize 14 evaluation metrics under four categories for different outputs of LJP tasks; (3) we review 12 legal-domain pretrained models in 3 languages and highlight 3 major research directions for LJP; (4) we show the state-of-art results for 8 representative datasets from different court cases and discuss the open challenges. This paper can provide up-to-date and comprehensive reviews to help readers understand the status of LJP. We hope to facilitate both NLP researchers and legal professionals for further joint efforts in this problem.

preprint2022arXiv

AutoIP: A United Framework to Integrate Physics into Gaussian Processes

Physical modeling is critical for many modern science and engineering applications. From a data science or machine learning perspective, where more domain-agnostic, data-driven models are pervasive, physical knowledge -- often expressed as differential equations -- is valuable in that it is complementary to data, and it can potentially help overcome issues such as data sparsity, noise, and inaccuracy. In this work, we propose a simple, yet powerful and general framework -- AutoIP, for Automatically Incorporating Physics -- that can integrate all kinds of differential equations into Gaussian Processes (GPs) to enhance prediction accuracy and uncertainty quantification. These equations can be linear or nonlinear, spatial, temporal, or spatio-temporal, complete or incomplete with unknown source terms, and so on. Based on kernel differentiation, we construct a GP prior to sample the values of the target function, equation-related derivatives, and latent source functions, which are all jointly from a multivariate Gaussian distribution. The sampled values are fed to two likelihoods: one to fit the observations, and the other to conform to the equation. We use the whitening method to evade the strong dependency between the sampled function values and kernel parameters, and we develop a stochastic variational learning algorithm. AutoIP shows improvement upon vanilla GPs in both simulation and several real-world applications, even using rough, incomplete equations.

preprint2022arXiv

Balanced Contrastive Learning for Long-Tailed Visual Recognition

Real-world data typically follow a long-tailed distribution, where a few majority categories occupy most of the data while most minority categories contain a limited number of samples. Classification models minimizing cross-entropy struggle to represent and classify the tail classes. Although the problem of learning unbiased classifiers has been well studied, methods for representing imbalanced data are under-explored. In this paper, we focus on representation learning for imbalanced data. Recently, supervised contrastive learning has shown promising performance on balanced data recently. However, through our theoretical analysis, we find that for long-tailed data, it fails to form a regular simplex which is an ideal geometric configuration for representation learning. To correct the optimization behavior of SCL and further improve the performance of long-tailed visual recognition, we propose a novel loss for balanced contrastive learning (BCL). Compared with SCL, we have two improvements in BCL: class-averaging, which balances the gradient contribution of negative classes; class-complement, which allows all classes to appear in every mini-batch. The proposed balanced contrastive learning (BCL) method satisfies the condition of forming a regular simplex and assists the optimization of cross-entropy. Equipped with BCL, the proposed two-branch framework can obtain a stronger feature representation and achieve competitive performance on long-tailed benchmark datasets such as CIFAR-10-LT, CIFAR-100-LT, ImageNet-LT, and iNaturalist2018. Our code is available at https://github.com/FlamieZhu/BCL .

preprint2022arXiv

Deep Quality Assessment of Compressed Videos: A Subjective and Objective Study

In the video coding process, the perceived quality of a compressed video is evaluated by full-reference quality evaluation metrics. However, it is difficult to obtain reference videos with perfect quality. To solve this problem, it is critical to design no-reference compressed video quality assessment algorithms, which assists in measuring the quality of experience on the server side and resource allocation on the network side. Convolutional Neural Network (CNN) has shown its advantage in Video Quality Assessment (VQA) with promising successes in recent years. A large-scale quality database is very important for learning accurate and powerful compressed video quality metrics. In this work, a semi-automatic labeling method is adopted to build a large-scale compressed video quality database, which allows us to label a large number of compressed videos with manageable human workload. The resulting Compressed Video quality database with Semi-Automatic Ratings (CVSAR), so far the largest of compressed video quality database. We train a no-reference compressed video quality assessment model with a 3D CNN for SpatioTemporal Feature Extraction and Evaluation (STFEE). Experimental results demonstrate that the proposed method outperforms state-of-the-art metrics and achieves promising generalization performance in cross-database tests. The CVSAR database and STFEE model will be made publicly available to facilitate reproducible research.

preprint2022arXiv

Detecting Algorithmically Generated Domains Using a GCNN-LSTM Hybrid Neural Network

Domain generation algorithm (DGA) is used by botnets to build a stealthy command and control (C&C) communication channel between the C&C server and the bots. A DGA can periodically produce a large number of pseudo-random algorithmically generated domains (AGDs). AGD detection algorithms provide a lightweight, promising solution in response to the existing DGA techniques. In this paper, a GCNN (gated convolutional neural network)-LSTM (long short-term memory) Hybrid Neural Network (GLHNN) for AGD detection is proposed. In GLHNN, GCNN is applied to extract the informative features from domain names on top of LSTM which further processes the feature sequence. GLHNN is experimentally validated using representative AGDs covering six classes of DGAs. GLHNN is compared with the state-of-the-art detection models and demonstrates the best overall detection performance among these tested models.

preprint2022arXiv

Dynamic GPU Energy Optimization for Machine Learning Training Workloads

GPUs are widely used to accelerate the training of machine learning workloads. As modern machine learning models become increasingly larger, they require a longer time to train, leading to higher GPU energy consumption. This paper presents GPOEO, an online GPU energy optimization framework for machine learning training workloads. GPOEO dynamically determines the optimal energy configuration by employing novel techniques for online measurement, multi-objective prediction modeling, and search optimization. To characterize the target workload behavior, GPOEO utilizes GPU performance counters. To reduce the performance counter profiling overhead, it uses an analytical model to detect the training iteration change and only collects performance counter data when an iteration shift is detected. GPOEO employs multi-objective models based on gradient boosting and a local search algorithm to find a trade-off between execution time and energy consumption. We evaluate the GPOEO by applying it to 71 machine learning workloads from two AI benchmark suites running on an NVIDIA RTX3080Ti GPU. Compared with the NVIDIA default scheduling strategy, GPOEO delivers a mean energy saving of 16.2% with a modest average execution time increase of 5.1%.

preprint2022arXiv

Exploring the Impact of Negative Samples of Contrastive Learning: A Case Study of Sentence Embedding

Contrastive learning is emerging as a powerful technique for extracting knowledge from unlabeled data. This technique requires a balanced mixture of two ingredients: positive (similar) and negative (dissimilar) samples. This is typically achieved by maintaining a queue of negative samples during training. Prior works in the area typically uses a fixed-length negative sample queue, but how the negative sample size affects the model performance remains unclear. The opaque impact of the number of negative samples on performance when employing contrastive learning aroused our in-depth exploration. This paper presents a momentum contrastive learning model with negative sample queue for sentence embedding, namely MoCoSE. We add the prediction layer to the online branch to make the model asymmetric and together with EMA update mechanism of the target branch to prevent the model from collapsing. We define a maximum traceable distance metric, through which we learn to what extent the text contrastive learning benefits from the historical information of negative samples. Our experiments find that the best results are obtained when the maximum traceable distance is at a certain range, demonstrating that there is an optimal range of historical information for a negative sample queue. We evaluate the proposed unsupervised MoCoSE on the semantic text similarity (STS) task and obtain an average Spearman's correlation of $77.27\%$. Source code is available at https://github.com/xbdxwyh/mocose.

preprint2022arXiv

Geo-Localization via Ground-to-Satellite Cross-View Image Retrieval

The large variation of viewpoint and irrelevant content around the target always hinder accurate image retrieval and its subsequent tasks. In this paper, we investigate an extremely challenging task: given a ground-view image of a landmark, we aim to achieve cross-view geo-localization by searching out its corresponding satellite-view images. Specifically, the challenge comes from the gap between ground-view and satellite-view, which includes not only large viewpoint changes (some parts of the landmark may be invisible from front view to top view) but also highly irrelevant background (the target landmark tend to be hidden in other surrounding buildings), making it difficult to learn a common representation or a suitable mapping. To address this issue, we take advantage of drone-view information as a bridge between ground-view and satellite-view domains. We propose a Peer Learning and Cross Diffusion (PLCD) framework. PLCD consists of three parts: 1) a peer learning across ground-view and drone-view to find visible parts to benefit ground-drone cross-view representation learning; 2) a patch-based network for satellite-drone cross-view representation learning; 3) a cross diffusion between ground-drone space and satellite-drone space. Extensive experiments conducted on the University-Earth and University-Google datasets show that our method outperforms state-of-the-arts significantly.

preprint2022arXiv

GMI-DRL: Empowering Multi-GPU Deep Reinforcement Learning with GPU Spatial Multiplexing

With the increasing popularity of robotics in industrial control and autonomous driving, deep reinforcement learning (DRL) raises the attention of various fields. However, DRL computation on the modern powerful GPU platform is still inefficient due to its heterogeneous workloads and interleaved execution paradigm. To this end, we propose GMI-DRL, a systematic design to accelerate multi-GPU DRL via GPU spatial multiplexing. We introduce a novel design of resource-adjustable GPU multiplexing instances (GMIs) to match the actual needs of DRL tasks, an adaptive GMI management strategy to simultaneously achieve high GPU utilization and computation throughput, and a highly efficient inter-GMI communication support to meet the demands of various DRL communication patterns. Comprehensive experiments reveal that GMI-DRL outperforms state-of-the-art NVIDIA Isaac Gym with NCCL (up to 2.81X) and Horovod (up to 2.34X) support in training throughput on the latest DGX-A100 platform. Our work provides an initial user experience with GPU spatial multiplexing in processing heterogeneous workloads with a mixture of computation and communication.

preprint2022arXiv

Improving Generalization of Metric Learning via Listwise Self-distillation

Most deep metric learning (DML) methods employ a strategy that forces all positive samples to be close in the embedding space while keeping them away from negative ones. However, such a strategy ignores the internal relationships of positive (negative) samples and often leads to overfitting, especially in the presence of hard samples and mislabeled samples. In this work, we propose a simple yet effective regularization, namely Listwise Self-Distillation (LSD), which progressively distills a model's own knowledge to adaptively assign a more appropriate distance target to each sample pair in a batch. LSD encourages smoother embeddings and information mining within positive (negative) samples as a way to mitigate overfitting and thus improve generalization. Our LSD can be directly integrated into general DML frameworks. Extensive experiments show that LSD consistently boosts the performance of various metric learning methods on multiple datasets.

preprint2022arXiv

Magic ELF: Image Deraining Meets Association Learning and Transformer

Convolutional neural network (CNN) and Transformer have achieved great success in multimedia applications. However, little effort has been made to effectively and efficiently harmonize these two architectures to satisfy image deraining. This paper aims to unify these two architectures to take advantage of their learning merits for image deraining. In particular, the local connectivity and translation equivariance of CNN and the global aggregation ability of self-attention (SA) in Transformer are fully exploited for specific local context and global structure representations. Based on the observation that rain distribution reveals the degradation location and degree, we introduce degradation prior to help background recovery and accordingly present the association refinement deraining scheme. A novel multi-input attention module (MAM) is proposed to associate rain perturbation removal and background recovery. Moreover, we equip our model with effective depth-wise separable convolutions to learn the specific feature representations and trade off computational complexity. Extensive experiments show that our proposed method (dubbed as ELF) outperforms the state-of-the-art approach (MPRNet) by 0.25 dB on average, but only accounts for 11.7\% and 42.1\% of its computational cost and parameters. The source code is available at https://github.com/kuijiang94/Magic-ELF.

preprint2022arXiv

Mass Testing and Characterization of 20-inch PMTs for JUNO

Main goal of the JUNO experiment is to determine the neutrino mass ordering using a 20kt liquid-scintillator detector. Its key feature is an excellent energy resolution of at least 3 % at 1 MeV, for which its instruments need to meet a certain quality and thus have to be fully characterized. More than 20,000 20-inch PMTs have been received and assessed by JUNO after a detailed testing program which began in 2017 and elapsed for about four years. Based on this mass characterization and a set of specific requirements, a good quality of all accepted PMTs could be ascertained. This paper presents the performed testing procedure with the designed testing systems as well as the statistical characteristics of all 20-inch PMTs intended to be used in the JUNO experiment, covering more than fifteen performance parameters including the photocathode uniformity. This constitutes the largest sample of 20-inch PMTs ever produced and studied in detail to date, i.e. 15,000 of the newly developed 20-inch MCP-PMTs from Northern Night Vision Technology Co. (NNVT) and 5,000 of dynode PMTs from Hamamatsu Photonics K. K.(HPK).

preprint2022arXiv

Microwave heating effect on diamond sample of NV centers

Diamond samples of defects with negative charged nitrogen-vacancy (NV) centers are promising solid state spin sensors suitable for quantum information processing, high sensitive measurements of magnetic, electric and thermal fields in nanoscale. The diamond defect with a NV center is unique for its robust temperature-dependent zero field splitting Dgs of the triplet ground state. This property enables optical readout of electron spin states through manipulation of the ground triplet state using microwave resonance with Dgs from 100 K to about 600 K. Thus, prohibiting Dgs from unwanted external thermal disturbances is crucial for an accurate measurement using diamond NV sensors. Our observation demonstrates the existence of a prominent microwave heating effect on the diamond samples of NV centers. The effect is inevitable to shift Dgs and cause measurement errors. The temperature increment caused by the effect monotonically depends on the power and the duration of microwave irradiation. The effect is obvious with the microwave irradiation in the continuous mode and some pulse sequence modes, but is neglectable for the quantum lock-in XY8-N method.

preprint2022arXiv

Nonparametric Embeddings of Sparse High-Order Interaction Events

High-order interaction events are common in real-world applications. Learning embeddings that encode the complex relationships of the participants from these events is of great importance in knowledge mining and predictive tasks. Despite the success of existing approaches, e.g. Poisson tensor factorization, they ignore the sparse structure underlying the data, namely the occurred interactions are far less than the possible interactions among all the participants. In this paper, we propose Nonparametric Embeddings of Sparse High-order interaction events (NESH). We hybridize a sparse hypergraph (tensor) process and a matrix Gaussian process to capture both the asymptotic structural sparsity within the interactions and nonlinear temporal relationships between the participants. We prove strong asymptotic bounds (including both a lower and an upper bound) of the sparsity ratio, which reveals the asymptotic properties of the sampled structure. We use batch-normalization, stick-breaking construction, and sparse variational GP approximations to develop an efficient, scalable model inference algorithm. We demonstrate the advantage of our approach in several real-world applications.

preprint2022arXiv

Nonparametric Factor Trajectory Learning for Dynamic Tensor Decomposition

Tensor decomposition is a fundamental framework to analyze data that can be represented by multi-dimensional arrays. In practice, tensor data is often accompanied by temporal information, namely the time points when the entry values were generated. This information implies abundant, complex temporal variation patterns. However, current methods always assume the factor representations of the entities in each tensor mode are static, and never consider their temporal evolution. To fill this gap, we propose NONparametric FActor Trajectory learning for dynamic tensor decomposition (NONFAT). We place Gaussian process (GP) priors in the frequency domain and conduct inverse Fourier transform via Gauss-Laguerre quadrature to sample the trajectory functions. In this way, we can overcome data sparsity and obtain robust trajectory estimates across long time horizons. Given the trajectory values at specific time points, we use a second-level GP to sample the entry values and to capture the temporal relationship between the entities. For efficient and scalable inference, we leverage the matrix Gaussian structure in the model, introduce a matrix Gaussian posterior, and develop a nested sparse variational learning algorithm. We have shown the advantage of our method in several real-world applications.

preprint2022arXiv

Physics Informed Deep Kernel Learning

Deep kernel learning is a promising combination of deep neural networks and nonparametric function learning. However, as a data driven approach, the performance of deep kernel learning can still be restricted by scarce or insufficient data, especially in extrapolation tasks. To address these limitations, we propose Physics Informed Deep Kernel Learning (PI-DKL) that exploits physics knowledge represented by differential equations with latent sources. Specifically, we use the posterior function sample of the Gaussian process as the surrogate for the solution of the differential equation, and construct a generative component to integrate the equation in a principled Bayesian hybrid framework. For efficient and effective inference, we marginalize out the latent variables in the joint probability and derive a collapsed model evidence lower bound (ELBO), based on which we develop a stochastic model estimation algorithm. Our ELBO can be viewed as a nice, interpretable posterior regularization objective. On synthetic datasets and real-world applications, we show the advantage of our approach in both prediction accuracy and uncertainty quantification.

preprint2022arXiv

Reference-Guided Texture and Structure Inference for Image Inpainting

Existing learning-based image inpainting methods are still in challenge when facing complex semantic environments and diverse hole patterns. The prior information learned from the large scale training data is still insufficient for these situations. Reference images captured covering the same scenes share similar texture and structure priors with the corrupted images, which offers new prospects for the image inpainting tasks. Inspired by this, we first build a benchmark dataset containing 10K pairs of input and reference images for reference-guided inpainting. Then we adopt an encoder-decoder structure to separately infer the texture and structure features of the input image considering their pattern discrepancy of texture and structure during inpainting. A feature alignment module is further designed to refine these features of the input image with the guidance of a reference image. Both quantitative and qualitative evaluations demonstrate the superiority of our method over the state-of-the-art methods in terms of completing complex holes.

preprint2022arXiv

Relaxation Oscillations of an Exciton-polariton Condensate Driven by Parametric Scattering

We report observation of coherent oscillations in the relaxation dynamics of an exciton-polariton condensate driven by parametric scattering processes. As a result of the interbranch scattering scheme and the nonlinear polariton-polariton interactions, such parametric scatterings exhibit high scattering efficiency, which leads to fast depletion of the polariton condensate and periodic shut-off of the bosonic stimulation processes, eventually causing relaxation oscillations. Employing polariton-reservoir interactions, the oscillation dynamics in the time domain can be projected onto the energy space. In theory, our simulations using the open-dissipative Gross-Pitaevskii equation are in excellent agreement with experimental observations. Surprisingly, the oscillation patterns are clearly visible in our time-integrated images including many excitation pulses, implying the high stability of the relaxation oscillations driven by polariton parametric scatterings.

preprint2022arXiv

Robustness of Neural Architectures for Audio Event Detection

Traditionally, in Audio Recognition pipeline, noise is suppressed by the "frontend", relying on preprocessing techniques such as speech enhancement. However, it is not guaranteed that noise will not cascade into downstream pipelines. To understand the actual influence of noise on the entire audio pipeline, in this paper, we directly investigate the impact of noise on a different types of neural models without the preprocessing step. We measure the recognition performances of 4 different neural network models on the task of environment sound classification under the 3 types of noises: \emph{occlusion} (to emulate intermittent noise), \emph{Gaussian} noise (models continuous noise), and \emph{adversarial perturbations} (worst case scenario). Our intuition is that the different ways in which these models process their input (i.e. CNNs have strong locality inductive biases, which Transformers do not have) should lead to observable differences in performance and/ or robustness, an understanding of which will enable further improvements. We perform extensive experiments on AudioSet which is the largest weakly-labeled sound event dataset available. We also seek to explain the behaviors of different models through output distribution change and weight visualization.

preprint2022arXiv

Spatial-Temporal Space Hand-in-Hand: Spatial-Temporal Video Super-Resolution via Cycle-Projected Mutual Learning

Spatial-Temporal Video Super-Resolution (ST-VSR) aims to generate super-resolved videos with higher resolution(HR) and higher frame rate (HFR). Quite intuitively, pioneering two-stage based methods complete ST-VSR by directly combining two sub-tasks: Spatial Video Super-Resolution (S-VSR) and Temporal Video Super-Resolution(T-VSR) but ignore the reciprocal relations among them. Specifically, 1) T-VSR to S-VSR: temporal correlations help accurate spatial detail representation with more clues; 2) S-VSR to T-VSR: abundant spatial information contributes to the refinement of temporal prediction. To this end, we propose a one-stage based Cycle-projected Mutual learning network (CycMu-Net) for ST-VSR, which makes full use of spatial-temporal correlations via the mutual learning between S-VSR and T-VSR. Specifically, we propose to exploit the mutual information among them via iterative up-and-down projections, where the spatial and temporal features are fully fused and distilled, helping the high-quality video reconstruction. Besides extensive experiments on benchmark datasets, we also compare our proposed CycMu-Net with S-VSR and T-VSR tasks, demonstrating that our method significantly outperforms state-of-the-art methods.

preprint2022arXiv

Towards Generalizable Person Re-identification with a Bi-stream Generative Model

Generalizable person re-identification (re-ID) has attracted growing attention due to its powerful adaptation capability in the unseen data domain. However, existing solutions often neglect either crossing cameras (e.g., illumination and resolution differences) or pedestrian misalignments (e.g., viewpoint and pose discrepancies), which easily leads to poor generalization capability when adapted to the new domain. In this paper, we formulate these difficulties as: 1) Camera-Camera (CC) problem, which denotes the various human appearance changes caused by different cameras; 2) Camera-Person (CP) problem, which indicates the pedestrian misalignments caused by the same identity person under different camera viewpoints or changing pose. To solve the above issues, we propose a Bi-stream Generative Model (BGM) to learn the fine-grained representations fused with camera-invariant global feature and pedestrian-aligned local feature, which contains an encoding network and two stream decoding sub-networks. Guided by original pedestrian images, one stream is employed to learn a camera-invariant global feature for the CC problem via filtering cross-camera interference factors. For the CP problem, another stream learns a pedestrian-aligned local feature for pedestrian alignment using information-complete densely semantically aligned part maps. Moreover, a part-weighted loss function is presented to reduce the influence of missing parts on pedestrian alignment. Extensive experiments demonstrate that our method outperforms the state-of-the-art methods on the large-scale generalizable re-ID benchmarks, involving domain generalization setting and cross-domain setting.

preprint2022arXiv

Understanding Adversarial Robustness of Vision Transformers via Cauchy Problem

Recent research on the robustness of deep learning has shown that Vision Transformers (ViTs) surpass the Convolutional Neural Networks (CNNs) under some perturbations, e.g., natural corruption, adversarial attacks, etc. Some papers argue that the superior robustness of ViT comes from the segmentation of its input images; others say that the Multi-head Self-Attention (MSA) is the key to preserving the robustness. In this paper, we aim to introduce a principled and unified theoretical framework to investigate such an argument on ViT's robustness. We first theoretically prove that, unlike Transformers in Natural Language Processing, ViTs are Lipschitz continuous. Then we theoretically analyze the adversarial robustness of ViTs from the perspective of the Cauchy Problem, via which we can quantify how the robustness propagates through layers. We demonstrate that the first and last layers are the critical factors to affect the robustness of ViTs. Furthermore, based on our theory, we empirically show that unlike the claims from existing research, MSA only contributes to the adversarial robustness of ViTs under weak adversarial attacks, e.g., FGSM, and surprisingly, MSA actually comprises the model's adversarial robustness under stronger attacks, e.g., PGD attacks.

preprint2022arXiv

Unsupervised Foggy Scene Understanding via Self Spatial-Temporal Label Diffusion

Understanding foggy image sequence in the driving scenes is critical for autonomous driving, but it remains a challenging task due to the difficulty in collecting and annotating real-world images of adverse weather. Recently, the self-training strategy has been considered a powerful solution for unsupervised domain adaptation, which iteratively adapts the model from the source domain to the target domain by generating target pseudo labels and re-training the model. However, the selection of confident pseudo labels inevitably suffers from the conflict between sparsity and accuracy, both of which will lead to suboptimal models. To tackle this problem, we exploit the characteristics of the foggy image sequence of driving scenes to densify the confident pseudo labels. Specifically, based on the two discoveries of local spatial similarity and adjacent temporal correspondence of the sequential image data, we propose a novel Target-Domain driven pseudo label Diffusion (TDo-Dif) scheme. It employs superpixels and optical flows to identify the spatial similarity and temporal correspondence, respectively and then diffuses the confident but sparse pseudo labels within a superpixel or a temporal corresponding pair linked by the flow. Moreover, to ensure the feature similarity of the diffused pixels, we introduce local spatial similarity loss and temporal contrastive loss in the model re-training stage. Experimental results show that our TDo-Dif scheme helps the adaptive model achieve 51.92% and 53.84% mean intersection-over-union (mIoU) on two publicly available natural foggy datasets (Foggy Zurich and Foggy Driving), which exceeds the state-of-the-art unsupervised domain adaptive semantic segmentation methods. Models and data can be found at https://github.com/velor2012/TDo-Dif.

preprint2022arXiv

Unsupervised Manga Character Re-identification via Face-body and Spatial-temporal Associated Clustering

In the past few years, there has been a dramatic growth in e-manga (electronic Japanese-style comics). Faced with the booming demand for manga research and the large amount of unlabeled manga data, we raised a new task, called unsupervised manga character re-identification. However, the artistic expression and stylistic limitations of manga pose many challenges to the re-identification problem. Inspired by the idea that some content-related features may help clustering, we propose a Face-body and Spatial-temporal Associated Clustering method (FSAC). In the face-body combination module, a face-body graph is constructed to solve problems such as exaggeration and deformation in artistic creation by using the integrity of the image. In the spatial-temporal relationship correction module, we analyze the appearance features of characters and design a temporal-spatial-related triplet loss to fine-tune the clustering. Extensive experiments on a manga book dataset with 109 volumes validate the superiority of our method in unsupervised manga character re-identification.

preprint2022arXiv

Visual-Tactile Sensing for Real-time Liquid Volume Estimation in Grasping

We propose a deep visuo-tactile model for realtime estimation of the liquid inside a deformable container in a proprioceptive way.We fuse two sensory modalities, i.e., the raw visual inputs from the RGB camera and the tactile cues from our specific tactile sensor without any extra sensor calibrations.The robotic system is well controlled and adjusted based on the estimation model in real time. The main contributions and novelties of our work are listed as follows: 1) Explore a proprioceptive way for liquid volume estimation by developing an end-to-end predictive model with multi-modal convolutional networks, which achieve a high precision with an error of around 2 ml in the experimental validation. 2) Propose a multi-task learning architecture which comprehensively considers the losses from both classification and regression tasks, and comparatively evaluate the performance of each variant on the collected data and actual robotic platform. 3) Utilize the proprioceptive robotic system to accurately serve and control the requested volume of liquid, which is continuously flowing into a deformable container in real time. 4) Adaptively adjust the grasping plan to achieve more stable grasping and manipulation according to the real-time liquid volume prediction.

preprint2022arXiv

You Only Align Once: Bidirectional Interaction for Spatial-Temporal Video Super-Resolution

Spatial-Temporal Video Super-Resolution (ST-VSR) technology generates high-quality videos with higher resolution and higher frame rates. Existing advanced methods accomplish ST-VSR tasks through the association of Spatial and Temporal video super-resolution (S-VSR and T-VSR). These methods require two alignments and fusions in S-VSR and T-VSR, which is obviously redundant and fails to sufficiently explore the information flow of consecutive spatial LR frames. Although bidirectional learning (future-to-past and past-to-future) was introduced to cover all input frames, the direct fusion of final predictions fails to sufficiently exploit intrinsic correlations of bidirectional motion learning and spatial information from all frames. We propose an effective yet efficient recurrent network with bidirectional interaction for ST-VSR, where only one alignment and fusion is needed. Specifically, it first performs backward inference from future to past, and then follows forward inference to super-resolve intermediate frames. The backward and forward inferences are assigned to learn structures and details to simplify the learning task with joint optimizations. Furthermore, a Hybrid Fusion Module (HFM) is designed to aggregate and distill information to refine spatial information and reconstruct high-quality video frames. Extensive experiments on two public datasets demonstrate that our method outperforms state-of-the-art methods in efficiency, and reduces calculation cost by about 22%.

preprint2021arXiv

Antiferroelectric negative capacitance from a structural phase transition in zirconia

Crystalline materials with broken inversion symmetry can exhibit a spontaneous electric polarization, which originates from a microscopic electric dipole moment. Long-range polar or anti-polar order of such permanent dipoles gives rise to ferroelectricity or antiferroelectricity, respectively. However, the recently discovered antiferroelectrics of fluorite structure (HfO$_2$ and ZrO$_2$) are different: A non-polar phase transforms into a polar phase by spontaneous inversion symmetry breaking upon the application of an electric field. Here, we show that this structural transition in antiferroelectric ZrO$_2$ gives rise to a negative capacitance, which is promising for overcoming the fundamental limits of energy efficiency in electronics. Our findings provide insight into the thermodynamically 'forbidden' region of the antiferroelectric transition in ZrO$_2$ and extend the concept of negative capacitance beyond ferroelectricity. This shows that negative capacitance is a more general phenomenon than previously thought and can be expected in a much broader range of materials exhibiting structural phase transitions.

preprint2021arXiv

Interaction-aware Kalman Neural Networks for Trajectory Prediction

Forecasting the motion of surrounding obstacles (vehicles, bicycles, pedestrians and etc.) benefits the on-road motion planning for intelligent and autonomous vehicles. Complex scenes always yield great challenges in modeling the patterns of surrounding traffic. For example, one main challenge comes from the intractable interaction effects in a complex traffic system. In this paper, we propose a multi-layer architecture Interaction-aware Kalman Neural Networks (IaKNN) which involves an interaction layer for resolving high-dimensional traffic environmental observations as interaction-aware accelerations, a motion layer for transforming the accelerations to interaction aware trajectories, and a filter layer for estimating future trajectories with a Kalman filter network. Attributed to the multiple traffic data sources, our end-to-end trainable approach technically fuses dynamic and interaction-aware trajectories boosting the prediction performance. Experiments on the NGSIM dataset demonstrate that IaKNN outperforms the state-of-the-art methods in terms of effectiveness for traffic trajectory prediction.

preprint2021arXiv

JUNO Physics and Detector

The Jiangmen Underground Neutrino Observatory (JUNO) is a 20 kton LS detector at 700-m underground. An excellent energy resolution and a large fiducial volume offer exciting opportunities for addressing many important topics in neutrino and astro-particle physics. With 6 years of data, the neutrino mass ordering can be determined at 3-4 sigma and three oscillation parameters can be measured to a precision of 0.6% or better by detecting reactor antineutrinos. With 10 years of data, DSNB could be observed at 3-sigma; a lower limit of the proton lifetime of 8.34e33 years (90% C.L.) can be set by searching for p->nu_bar K^+; detection of solar neutrinos would shed new light on the solar metallicity problem and examine the vacuum-matter transition region. A core-collapse supernova at 10 kpc would lead to ~5000 IBD and ~2000 (300) all-flavor neutrino-proton (electron) scattering events. Geo-neutrinos can be detected with a rate of ~400 events/year. We also summarize the final design of the JUNO detector and the key R&D achievements. All 20-inch PMTs have been tested. The average photon detection efficiency is 28.9% for the 15,000 MCP PMTs and 28.1% for the 5,000 dynode PMTs, higher than the JUNO requirement of 27%. Together with the >20 m attenuation length of LS, we expect a yield of 1345 p.e. per MeV and an effective energy resolution of 3.02%/\sqrt{E (MeV)}$ in simulations. The underwater electronics is designed to have a loss rate <0.5% in 6 years. With degassing membranes and a micro-bubble system, the radon concentration in the 35-kton water pool could be lowered to <10 mBq/m^3. Acrylic panels of radiopurity <0.5 ppt U/Th are produced. The 20-kton LS will be purified onsite. Singles in the fiducial volume can be controlled to ~10 Hz. The JUNO experiment also features a double calorimeter system with 25,600 3-inch PMTs, a LS testing facility OSIRIS, and a near detector TAO.

preprint2021arXiv

Model Rectification via Unknown Unknowns Extraction from Deployment Samples

Model deficiency that results from incomplete training data is a form of structural blindness that leads to costly errors, oftentimes with high confidence. During the training of classification tasks, underrepresented class-conditional distributions that a given hypothesis space can recognize results in a mismatch between the model and the target space. To mitigate the consequences of this discrepancy, we propose Random Test Sampling and Cross-Validation (RTSCV) as a general algorithmic framework that aims to perform a post-training model rectification at deployment time in a supervised way. RTSCV extracts unknown unknowns (u.u.s), i.e., examples from the class-conditional distributions that a classifier is oblivious to, and works in combination with a diverse family of modern prediction models. RTSCV augments the training set with a sample of the test set (or deployment data) and uses this redefined class layout to discover u.u.s via cross-validation, without relying on active learning or budgeted queries to an oracle. We contribute a theoretical analysis that establishes performance guarantees based on the design bases of modern classifiers. Our experimental evaluation demonstrates RTSCV's effectiveness, using 7 benchmark tabular and computer vision datasets, by reducing a performance gap as large as 41% from the respective pre-rectification models. Last we show that RTSCV consistently outperforms state-of-the-art approaches.

Zheng Wang

Quick read

Decide how to stay connected

How to connect with this researcher

Open a focused conversation when the fit is right

See the researcher in context

Building this graph slice

96 published item(s)

An Enigmatic PeVatron in an Area around HII Region G35.6$-$0.5

An Ultrahigh-energy $γ$-ray Bubble Powered by a Super PeVatron

Beyond Prompts: Space-Time Decoupling Control-Plane Jailbreaks in LLM Structured Output

Beyond Seen Bounds: Class-Centric Polarization for Single-Domain Generalized Deep Metric Learning

Constraining the Cosmic-ray Energy Based on Observations of Nearby Galaxy Clusters by LHAASO

Constraints on heavy decaying dark matter from 570 days of LHAASO observations

Deep view of Composite SNR CTA1 with LHAASO in $γ$-rays up to 300 TeV

Discovery of a new $γ$-ray source LHAASO J0341+5258 with emission up to 200TeV

Discovery of the Ultra-high energy gamma-ray source LHAASO J2108+5157

Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning

Energy calibration of LHAASO-KM2A using the cosmic ray Moon shadow

Energy-Dependent Shifts of Medium-Scale Anisotropies in Very-High-Energy Cosmic Rays Observed by LHAASO-KM2A

Evidence for particle acceleration approaching PeV energies in the W51 complex

Exploring Lorentz Invariance Violation from Ultra-high-energy Gamma Rays Observed by LHAASO

Extended Very-High-Energy Gamma-Ray Emission Surrounding PSR J0622 + 3749 Observed by LHAASO-KM2A

First Submillimeter Lights from Dome A: Tracing the Carbon Cycle in the Feedback of Massive Stars

From Context to Skills: Can Language Models Learn from Context Skillfully?

LHAASO Detection of Ultra-High-Energy Gamma-Ray Emission toward the Giant Molecular Clouds

Measurement of attenuation length of the muon content in extensive air showers from 0.3 to 30 PeV with LHAASO

Measurement of Very-high-energy Diffuse Gamma-ray Emissions from the Galactic Plane with LHAASO-WCDA

Modality-Aware Identity Construction and Counterfactual Structure Learning for ID-Free Multimodal Recommendation

Reflections and New Directions for Human-Centered Large Language Models

Reinforced Collaboration in Multi-Agent Flow Networks

SuperEar: Eavesdropping on Mobile Voice Calls via Stealthy Acoustic Metamaterials

The New Compiler Stack: A Survey on the Synergy of LLMs and Compilers

Transient Large-Scale Anisotropy in TeV Cosmic Rays due to an Interplanetary Coronal Mass Ejection

CREPES-X: Hierarchical Bearing-Distance-Inertial Direct Cooperative Relative Pose Estimation System

Observation of the $γ$-ray Emission from W43 with LHAASO

RAJ-PGA: Reasoning-Activated Jailbreak and Principle-Guided Alignment Framework for Large Reasoning Models

Study of Ultra-High-Energy Gamma-Ray Source 1LHAASO J0056+6346u and Its Possible Origins

Ultrahigh-Energy Gamma-ray Emission Associated with Black Hole-Jet Systems

Yggdrasil: Bridging Dynamic Speculation and Static Runtime for Latency-Optimal Tree-Based LLM Decoding

A Survey on Legal Judgment Prediction: Datasets, Metrics, Models and Challenges

AutoIP: A United Framework to Integrate Physics into Gaussian Processes

Balanced Contrastive Learning for Long-Tailed Visual Recognition

Deep Quality Assessment of Compressed Videos: A Subjective and Objective Study

Detecting Algorithmically Generated Domains Using a GCNN-LSTM Hybrid Neural Network

Dynamic GPU Energy Optimization for Machine Learning Training Workloads

Exploring the Impact of Negative Samples of Contrastive Learning: A Case Study of Sentence Embedding

Geo-Localization via Ground-to-Satellite Cross-View Image Retrieval

GMI-DRL: Empowering Multi-GPU Deep Reinforcement Learning with GPU Spatial Multiplexing

Improving Generalization of Metric Learning via Listwise Self-distillation

Magic ELF: Image Deraining Meets Association Learning and Transformer

Mass Testing and Characterization of 20-inch PMTs for JUNO

Microwave heating effect on diamond sample of NV centers

Nonparametric Embeddings of Sparse High-Order Interaction Events

Nonparametric Factor Trajectory Learning for Dynamic Tensor Decomposition

Physics Informed Deep Kernel Learning

Reference-Guided Texture and Structure Inference for Image Inpainting

Relaxation Oscillations of an Exciton-polariton Condensate Driven by Parametric Scattering

Robustness of Neural Architectures for Audio Event Detection

Spatial-Temporal Space Hand-in-Hand: Spatial-Temporal Video Super-Resolution via Cycle-Projected Mutual Learning

Towards Generalizable Person Re-identification with a Bi-stream Generative Model

Understanding Adversarial Robustness of Vision Transformers via Cauchy Problem

Unsupervised Foggy Scene Understanding via Self Spatial-Temporal Label Diffusion

Unsupervised Manga Character Re-identification via Face-body and Spatial-temporal Associated Clustering

Visual-Tactile Sensing for Real-time Liquid Volume Estimation in Grasping

You Only Align Once: Bidirectional Interaction for Spatial-Temporal Video Super-Resolution

Antiferroelectric negative capacitance from a structural phase transition in zirconia

Interaction-aware Kalman Neural Networks for Trajectory Prediction

JUNO Physics and Detector

Model Rectification via Unknown Unknowns Extraction from Deployment Samples

3D Spectrum Mapping Based on ROI-Driven UAV Deployment

Adaptive driver-automation shared steering control via forearm surface electromyography measurement

Analysis of Truck Driver Behavior to Design Different Lane Change Styles in Automated Driving

Beyond Intra-modality: A Survey of Heterogeneous Person Re-identification

Constructing Geographic and Long-term Temporal Graph for Traffic Forecasting

Cosmic muon flux measurement and tunnel overburden structure imaging

Curriculum Audiovisual Learning

DTDN: Dual-task De-raining Network

Efficient and Effective Similar Subtrajectory Search with Deep Reinforcement Learning

Exploring Image Enhancement for Salient Object Detection in Low Light Images