arXiv subjects

Jinhui Guo

Publications and source records attributed to Jinhui Guo.

18 recordsLinked to original sources

arXiv2026

DICS: Exploring Data Intrinsic Consistency for Visual Instruction Selection

Visual instruction tuning is crucial for advancing the vision-language alignment and instruction-following capabilities of Vision-Language Models (VLMs). However, identifying optimal subsets under a fixed ratio constraint from rapidly expanding datasets remains a significant bottleneck. While existing methods largely depend on distribution diversity or heuristic filtering, they often overlook the internal coherence within individual samples. To bridge this gap, we propose Data Intrinsic Consistency (DIC), a self-scoring metric designed to quantify the sample-level inter-component consistency. DIC consists of two modules: Visual Information Consistency (VIC), evaluating the alignment between visual content and instructions, and Response Information Consistency (RIC), assessing response coherence relative to the instruction. Building upon DIC, we introduce Data Intrinsic Consistency Selection (DICS), an adaptive data selection method that optimizes the trade-off between high intra-sample consistency and global distributional diversity under varying data budgets. Extensive experiments demonstrate that DICS consistently outperforms state-of-the-art methods across diverse dataset scales and model architectures, surpassing full-dataset fine-tuning while using only 25% of the LLaVA-1.5-665K data. We further curate DICS-6M, a 6M-sample multi-modal instruction corpus that enables the largest-scale visual instruction selection study to date; remarkably, DICS reaches 94.52\% of the official InternVL3-8B-Instruct performance using less than 25\% of its reported training data. Code can be seen at https://github.com/cqu-student/DICS

cs.CV↗

arXiv2026

An Axial $U_A(1)_{L_\mu-L_\tau}$: UV Completion and Experimental Searches

We propose an anomaly-free and renormalizable axial $U_A(1)_{L_\mu-L_\tau}$ model and study its experimental signatures for $A'$ masses from the MeV scale to the TeV scale. The opposite charges of the left- and right-handed charged leptons forbid the usual muon and tau Yukawa interactions. Their masses are instead generated by a singlet scalar and heavy vector-like leptons through a universal-seesaw mechanism. We focus on heavy vector-like leptons, small light--heavy mixing, and $m_s\gtrsim10~\mathrm{GeV}$. In this limit, the observables considered here depend mainly on $(m_{A'},g_X)$, while the other model parameters are restricted by mixing and perturbativity. We confront this benchmark with current experimental searches. For neutrino trident production, our finite-$m_\mu$ calculation shows that $A'$ modifies the axial weak coefficient, rather than the vector coefficient relevant to the usual $L_\mu-L_\tau$ model. The longitudinal mode enhances muon bremsstrahlung and gives a negative contribution to $(g-2)_\mu$; the latter dominates over the scalar contribution in our benchmark. Combining these results with invisible meson decays and four-muon resonance searches, we summarize the phenomenological constraints in the $(m_{A'},g_X)$ plane. For $m_{A'}\gg m_\mu$, vector and axial final-state-radiation rates become nearly identical, so the corresponding collider limits can be obtained by rate matching. At a future muon collider, the total rate alone does not fully resolve the interaction structure, whereas angular distributions, especially the forward--backward asymmetry in $\mu^+\mu^-\to\tau^+\tau^-$, retain direct sensitivity to chirality.

hep-ph↗

arXiv2026

Delayed Optimizer-State Transport Shapes Short-Horizon Training Decisions

Adaptive optimizers retain gradient history in moment variables, allowing a local change in loss weighting to alter later updates. We examine whether this delayed transport is large enough to change prospective short-horizon decisions. On committed future-minibatch sequences, we differentiate eight-step AdamW trajectories through the complete model--optimizer state and select exposure-matched Math--Code loss schedules before independent evaluation. Across 12 unused 0.3M Transformer histories, full transport lowers token-disjoint loss relative to an optimizer-aware immediate derivative in 10/12 histories (mean benefit $4.71\times10^{-4}$; exact one-sided sign test, $p=0.0193$). The two controllers act equally often but select different schedules in 60/96 windows. Crossed checkpoint--future-path tests attribute this reordering to the interaction between optimizer state and near-future data, while an independent Ising--CNN experiment shows that deleting moment-state transport destroys accurate response prediction. Full-transport scores also concentrate exact-rollout winners in larger candidate libraries, focusing finite-amplitude evaluation on a shortlist. On these committed short paths, optimizer memory and near-future data order are therefore actionable components of the training state, providing a mechanism-based criterion for when finite-horizon rather than one-step intervention is required.

cs.LG↗

arXiv2026

Down-Type Jet Identification in Fully Hadronic $t\bar t$ Events

Fully hadronic $t\bar t$ events carry the largest branching fraction and provide the down-type quark as a near-maximal spin analyzer ($\beta_d\simeq1$), yet they are rarely used for spin-correlation and entanglement studies because the down-type jets must be extracted from a difficult jet-to-parton reconstruction in the presence of a large QCD multijet background. We develop a two-stage machine-learning reconstruction for this channel: a GNN+Transformer network assigns all reconstructed jets into a legal six-jet top-pair candidate, and a second hybrid classifier resolves the four down/up hypotheses inside the two hadronic $W$ decays. The second stage combines an assignment scorer with an auxiliary conditional diffusion head; we find that the diffusion assignment score can rank hypotheses only when trained with a margin ranking objective, while standard denoising objectives leave it at the random baseline, and the two objectives are largely decoupled. We evaluate the method against a calibrated QCD background and an unmatched $t\bar t$ component using both reconstruction and spin observables. Compared with a ST1 only benchmark with random down/up labels, the learned ST2 classifier improves both the effective spin-analysis factor and the purity-weighted six-parton exact fraction at fixed event selection. A direct check of the standard spin-correlation coefficient $D$ further gives an ML-reconstructed value compatible with the truth-level value within the resampled spread. These results show that down/up identification can retain useful spin-correlation information in the all-hadronic channel.

hep-ph↗

arXiv2026

Atomic-scale sensing of photoexcitation processes in electronically isolated molecules via atomic force microscopy

Atomic-scale insights into light-matter interaction can be obtained using light-assisted scanning probe microscopy techniques. Recently, photoexcited charge carriers have been detected by means of scanning tunnelling microscopy, enabling the study of photo-induced charge transfer with atomic-scale spatial resolution. Here, we propose an approach based on atomic force microscopy (AFM), namely photoexcitation single-charge AFM (PE-AFM), to detect photoexcitation in single molecules adsorbed on non-conductive dielectric films. Synchronizing laser pulses to the oscillation of the tip of an AFM enables the detection of electron tunnelling events that follow photoexcitation. We demonstrate the PE-AFM technique for individual copper phthalocyanine molecules and achieve photoexcitation-driven AFM contrast with {\aa}ngstr\"om spatial resolution. The observed sub-molecular contrast suggests the involvement of a long-lived quadruplet excited state. When combined with the recently developed AFM excited-state spectroscopy including lifetime measurements, PE-AFM enables comprehensive characterization of electronic states involved in photoexcitation and subsequent intersystem crossing, establishing a powerful platform for investigating photophysical processes at the single-molecule level.

cond-mat.mes-hall↗

arXiv2026

Supermassive Primordial Black Holes from a Catalyzed Dark Phase Transition for Little Red Dots

JWST has revealed an abundant population of compact, low-metallicity "Little Red Dots" (LRDs) at high redshift, challenging conventional scenarios in which supermassive black holes (SMBHs) grow from stellar-mass seeds. We consider a scenario in which the SMBHs are instead supermassive primordial black holes (SMPBHs), formed directly in a decoupled, subdominant dark sector undergoing a first-order phase transition. Unlike conventional stochastic phase transitions, our mechanism is based on the catalysis by domain walls (DWs): most of the Universe completes the transition rapidly, while rare long-lived false-vacuum domains survive because of DW statistics and collapse into PBHs. This mechanism naturally yields SMPBH seeds with masses up to $M_{\rm PBH}\sim \mathcal{O}(10^{10}) M_\odot$, whose abundance can account for the observed LRD population. It also avoids the usual tensions with phase transition completion, $\Delta N_{\rm eff}$, and large curvature perturbations. The dark phase transition simultaneously generates an ultra-low-frequency stochastic gravitational-wave background peaking near the pulsar-timing-array range, providing a test of this dark-sector origin of LRDs.

hep-ph↗

arXiv2026

Ultralight Scalar Dark Matter with Off-Diagonal Flavor Couplings

Ultralight dark matter can behave as a coherent background field and induce time-dependent modifications of Standard Model parameters. We study a scenario in which a real ultralight scalar $\phi$ couples off-diagonally to down-type quarks, linking ultralight dark sectors to flavor physics. Working within an effective field theory, we diagonalize the quark mass matrix in a coherent $\phi$ background and derive analytic expressions for oscillatory shifts in down-type quark masses and CKM parameters. These effects lead to signatures in both the classical regime, where $\phi$ acts as a background field, and the quantum (particle) regime, where it contributes through on-shell production or off-shell mediation. Using precision flavor measurements, nuclear $\beta$ decays, atomic clocks, pulsar timing, and meson observables, we derive constraints on the flavor-violating couplings $\lambda_{ij}$ for $m_\phi \sim 10^{-24}$--$10^{-12}\,\mathrm{eV}$, highlighting the complementarity of time-domain and flavor probes of ultralight dark sectors.

hep-ph↗

arXiv2025

TrackGS: Optimizing COLMAP-Free 3D Gaussian Splatting with Global Track Constraints

We present TrackGS, a novel method to integrate global feature tracks with 3D Gaussian Splatting (3DGS) for COLMAP-free novel view synthesis. While 3DGS delivers impressive rendering quality, its reliance on accurate precomputed camera parameters remains a significant limitation. Existing COLMAP-free approaches depend on local constraints that fail in complex scenarios. Our key innovation lies in leveraging feature tracks to establish global geometric constraints, enabling simultaneous optimization of camera parameters and 3D Gaussians. Specifically, we: (1) introduce track-constrained Gaussians that serve as geometric anchors, (2) propose novel 2D and 3D track losses to enforce multi-view consistency, and (3) derive differentiable formulations for camera intrinsics optimization. Extensive experiments on challenging real-world and synthetic datasets demonstrate state-of-the-art performance, with much lower pose error than previous methods while maintaining superior rendering quality. Our approach eliminates the need for COLMAP preprocessing, making 3DGS more accessible for practical applications.

cs.CV↗

arXiv2025

NoPe-NeRF++: Local-to-Global Optimization of NeRF with No Pose Prior

In this paper, we introduce NoPe-NeRF++, a novel local-to-global optimization algorithm for training Neural Radiance Fields (NeRF) without requiring pose priors. Existing methods, particularly NoPe-NeRF, which focus solely on the local relationships within images, often struggle to recover accurate camera poses in complex scenarios. To overcome the challenges, our approach begins with a relative pose initialization with explicit feature matching, followed by a local joint optimization to enhance the pose estimation for training a more robust NeRF representation. This method significantly improves the quality of initial poses. Additionally, we introduce global optimization phase that incorporates geometric consistency constraints through bundle adjustment, which integrates feature trajectories to further refine poses and collectively boost the quality of NeRF. Notably, our method is the first work that seamlessly combines the local and global cues with NeRF, and outperforms state-of-the-art methods in both pose estimation accuracy and novel view synthesis. Extensive evaluations on benchmark datasets demonstrate our superior performance and robustness, even in challenging scenes, thus validating our design choices.

cs.CV↗

arXiv2025

Probing Scalar-Mediated Sterile Neutrinos with Gravitational Wave and Colliders Signals

We propose a UV-complete extension of the Standard Model in which a gauge-singlet scalar $S$ acquires a vacuum expectation value, generates a Majorana mass for a sterile neutrino $N$, and mixes with the Higgs field. This framework addresses neutrino masses via a seesaw mechanism and, for sufficiently large scalar mixing, can also drive a strong first-order electroweak phase transition, producing gravitational-wave (GW) signals potentially detectable by GW observatories. The Higgs-$S$ mixing also enhances sterile-neutrino pair production at colliders through $s$-channel exchange of the Higgs and $S$. Owing to the small active-sterile mixing angle, $N$ is generically long-lived, yielding characteristic displaced-vertex signatures. The combination of GW observations and displaced-vertex searches at colliders provides complementary cross-checks of the model parameter space.

hep-ph↗

arXiv2025

SD-VLM: Spatial Measuring and Understanding with Depth-Encoded Vision-Language Models

While vision language models (VLMs) excel in 2D semantic visual understanding, their ability to quantitatively reason about 3D spatial relationships remains under-explored, due to the deficiency of 2D images' spatial representation ability. In this paper, we analyze the problem hindering VLMs' spatial understanding abilities and propose SD-VLM, a novel framework that significantly enhances fundamental spatial perception abilities of VLMs through two key contributions: (1) propose Massive Spatial Measuring and Understanding (MSMU) dataset with precise spatial annotations, and (2) introduce a simple depth positional encoding method strengthening VLMs' spatial awareness. MSMU dataset covers massive quantitative spatial tasks with 700K QA pairs, 2.5M physical numerical annotations, and 10K chain-of-thought augmented samples. We have trained SD-VLM, a strong generalist VLM which shows superior quantitative spatial measuring and understanding capability. SD-VLM not only achieves state-of-the-art performance on our proposed MSMU-Bench, but also shows spatial generalization abilities on other spatial understanding benchmarks including Q-Spatial and SpatialRGPT-Bench. Extensive experiments demonstrate that SD-VLM outperforms GPT-4o and Intern-VL3-78B by 26.91% and 25.56% respectively on MSMU-Bench. Code and models are released at https://github.com/cpystan/SD-VLM.

cs.CV↗

arXiv2025

Probing Purely Inelastic Scalar Dark Matter Across Colliders and Gravitational Wave Observatories

We propose and study a purely inelastic scalar dark matter model, where two real scalars-dark matter $\phi_1$ and its excited partner $\phi_2$ interact with the Standard Model via a Higgs portal. After mass diagonalization, only inelastic couplings remain, allowing the model to evade stringent bounds from direct detection. We show that thermal (co-)annihilation between $\phi_1$ and $\phi_2$ naturally yields the observed dark matter relic abundance. The same interaction structure can induce a strongly first-order phase transition in the early universe, generating detectable gravitational waves in upcoming experiments. Meanwhile, the slight mass splitting between $\phi_1$ and $\phi_2$, along with the heavy off-shell mediator SM Higgs, leads to long-lived particle signatures of $\phi_2$ at the HL-LHC via the displaced muon-jets technique. We pinpoint a feasible parameter space where the correct relic abundance, observable gravitational waves, and collider signals can all be achieved concurrently, presenting a valuable chance to validate this scenario through a comprehensive examination encompassing cosmological, astrophysical, and collider investigations.

hep-ph↗

arXiv2025

Long-lived Sterile Neutrino Searches at Future Muon Colliders

We explore the potential of studying sterile neutrinos at a future high-energy muon collider, where these particles can generate small active neutrino masses via the seesaw mechanism and exhibit long-lived particle signatures. A Dirac sterile neutrino model with ${\rm U(1)}_{L_μ-L_τ}$ symmetry is introduced, where the heavy right-handed neutrino ($N_R$) produces tiny active neutrino masses, and the light left-handed neutrino ($N_L$) naturally behaves as a long-lived particle. The ${\rm U(1)}_{L_μ-L_τ}$ gauge symmetry also enhances sterile neutrino pair production at a future high-energy muon collider. Using the displaced vertex method, the muon collider can search for heavy sterile neutrino, especially for $m_L> m_W$. We find that a muon collider with $\sqrt{s} = 3~ (10)$ TeV and luminosity $L=1~(10)$ ab$^{-1}$ can probe $N_L$ masses of $m_L \in [100,~1500~(5000)]$ GeV and mixing angles $θ_{νL} \in [10^{-13},~10^{-6}]$.

hep-ph↗

arXiv2025

PTZ-Calib: Robust Pan-Tilt-Zoom Camera Calibration

In this paper, we present PTZ-Calib, a robust two-stage PTZ camera calibration method, that efficiently and accurately estimates camera parameters for arbitrary viewpoints. Our method includes an offline and an online stage. In the offline stage, we first uniformly select a set of reference images that sufficiently overlap to encompass a complete 360° view. We then utilize the novel PTZ-IBA (PTZ Incremental Bundle Adjustment) algorithm to automatically calibrate the cameras within a local coordinate system. Additionally, for practical application, we can further optimize camera parameters and align them with the geographic coordinate system using extra global reference 3D information. In the online stage, we formulate the calibration of any new viewpoints as a relocalization problem. Our approach balances the accuracy and computational efficiency to meet real-world demands. Extensive evaluations demonstrate our robustness and superior performance over state-of-the-art methods on various real and synthetic datasets. Datasets and source code can be accessed online at https://github.com/gjgjh/PTZ-Calib

cs.CV↗

arXiv2024

Long-lived Searches of Vector-like Lepton and Its Accompanying Scalar at Colliders

Recently, the vector-like leptons (VLLs) as a simple extension to the standard model (SM) have attracted widespread attention both in theory and experiments. The present collider searches mainly focus on the studies of their prompt decays, which prefer a relatively large coupling. In this paper, we concentrate on searches for long-lived signatures of the singlet VLLs $F$ or their accompanying scalar particles $ϕ$ both in the hadronic and electronic colliders. The long-lived signatures are naturally induced from small chiral mass mixing between VLLs and SM leptons. Two specific models distinguished by whether the VLLs couple to scalar particles are introduced to realize the aforementioned features. For long-lived VLLs case, we find that with the kink track method the sensitivities at future HL-LHC with $\sqrt{s}=14~\text{TeV}$ can reach the regions for VLL mass $m_F \in [200,1100]~\text{GeV}$ and the mass mixing parameter $θ_L \in [10^{-10},3\times 10^{-8}]$. For the long-lived accompanying scalar particle case, by fixing VLLs or scalar mass, or the mass ratio between VLL and the accompanying scalar, we explore the projected sensitivities through the time delay and displaced vertex strategies, which can probe the regions for $m_F \in [200,1200]~\text{GeV}$ and coupling $yθ_L\in [10^{-11},10^{-6}]$. Furthermore, we also explore the long-lived accompanying scalars at the future CEPC provided that the VLLs can couple to the SM first-generation leptons. We find that CEPC has good performances for $m_ϕ< 120~\text{GeV}$ and $m_F<1200~\text{GeV}$. These long-lived searches are complementary to previous studies, which opens the door towards the smaller coupling regions.

hep-ph↗

arXiv2024

New mechanism for primordial black hole formation from the QCD axion

We present a new mechanism for the primordial black hole (PBH) production within the QCD axion framework. We take the case where the Peccei-Quinn symmetry breaks during inflation, resulting in a $N_{\rm DW}=1$ string-wall network that re-enters horizon sufficiently late. Therefore, closed axion domain walls naturally arising in the network are sufficiently large to collapse into PBHs. Our numerical simulation shows that $\sim 0.3\%$ of the total wall area is in the form of closed walls. In addition, the relic abundance of dark matter is dominantly accounted for by free axions from the collapse of open walls bounded by strings. In this framework, the abundance of PBH within dark matter is calculated to be $\sim 0.9\%$. This fraction remains unaffected by axion parameters or the re-entering horizon temperature, as it is determined by the fixed proportion of closed walls in the network, governed by the principles of percolation theory. The resultant PBHs uniformly share the same mass, which spans from about $10^{-9}$ to $1$ solar mass, corresponding to the classical QCD axion mass window $10^{-5}-10^{-2}$~eV and the re-entering horizon temperature $300-1$~MeV. Intriguingly, PBHs in this mechanism can naturally account for the ultrashort-timescale gravitational microlensing events observed by the OGLE collaboration.

hep-ph↗

arXiv2023

Unveiling Time-Varying Signals of Ultralight Bosonic Dark Matter at Collider and Beam Dump Experiments

The ultralight boson represents a promising dark matter candidate exhibiting unique wave-like behaviors. These properties could transfer to the dark mediator, such as the kinetic mixing dark photon, which can be a link between the dark and Standard Model sectors, resulting in periodic oscillations of its mass. We propose a method to detect ultralight dark matter using dark mediators in collider and beam dump experiments, distinguishing it from conventional atomic, molecular, and optical methods. The time-varying nature of dark mediator mass exhibits a double-peak spectrum, reducing traditional constraints by 1 to 2 orders of magnitude, due to decreased luminosity exposure in each resonant mass bin. To enhance sensitivity, we utilize event time-stamps in the CMS Open Data and demonstrate that this technique boosts sensitivity by approximately one order of magnitude compared to the time-blind method. Moreover, it proves effective in detecting the invisible decay of the dark mediator.

hep-ph↗

arXiv2022

Heavy long-lived coannihilation partner from inelastic Dark Matter model and its signatures at the LHC

The coannihilation mechanism is a well-motivated alternative to the simple thermal freeze-out mechanism, where the dark matter relic density can be obtained through the coannihilation with a partner particle of similar mass with dark matter. When the partner particle is neutral, the inelastic nature of dark matter can help it to escape the direct detection limits. In this work, we focus on the coannihilation scenario in which the annihilation cross section is dominated by the partner-partner pair annihilation. We pay special interest on the parameter space where the coannihilation partner is long-lived, which leads to displaced signatures at the collider. In such case, it opens the heavy mass parameter space for the coannihilation dark matter, comparing with those dominated by the partner-dark matter annihilation. Specifically, we propose an inelastic scalar dark matter model with a broken symmetry, which realizes the domination of partner-partner pair annihilation. Then, we study two different realizations of the coannihilation partner decay and the existing constraints from the relic abundance, direct and indirect dark matter detection and the collider searches. We focus on the channel that the long-lived coannihilation partner decays to dark matter plus leptons. The high-luminosity LHC can reach good sensitivities for such heavy dark matter and coannihilation partner around 100--700 GeV.

hep-ph↗