arXiv subjects

Jing Gao

Publications and source records attributed to Jing Gao.

At least 55 records · Page 3Linked to original sources

arXiv2025

A 95 GeV Higgs boson and spontaneous CP-violation at the finite temperature

The ATLAS and CMS collaborations reported a diphoton excess in the invariant mass distribution around the 95.4 GeV with a local significance of $3.1σ$. Moreover, there is another $2.3σ$ local excess in the $b\bar{b}$ final state at LEP in the same mass region. A plausible solution is that the Higgs sector is extended to include an additional Higgs boson with a mass of $95.4$ GeV. We study a complex singlet scalar extension of the two-Higgs-doublet model in which the 95.4 GeV Higgs is from the mixing of three CP-even Higgs fields. In addition, the extended Higgs potential can achieve spontaneous CP-violation at the finite temperature and restore CP symmetry at the present temperature of the Universe. We find that the model can simultaneously explain the baryon asymmetry of the Universe, the diphoton and $b\bar{b}$ excesses around the 95.4 GeV while satisfying various relevant constraints including the experiments of collider and electric dipole moment.

hep-ph↗

arXiv2025

SUV: Scalable Large Language Model Copyright Compliance with Regularized Selective Unlearning

Large Language Models (LLMs) have transformed natural language processing by learning from massive datasets, yet this rapid progress has also drawn legal scrutiny, as the ability to unintentionally generate copyrighted content has already prompted several prominent lawsuits. In this work, we introduce SUV (Selective Unlearning for Verbatim data), a selective unlearning framework designed to prevent LLM from memorizing copyrighted content while preserving its overall utility. In detail, the proposed method constructs a dataset that captures instances of copyrighted infringement cases by the targeted LLM. With the dataset, we unlearn the content from the LLM by means of Direct Preference Optimization (DPO), which replaces the verbatim copyrighted content with plausible and coherent alternatives. Since DPO may hinder the LLM's performance in other unrelated tasks, we integrate gradient projection and Fisher information regularization to mitigate the degradation. We validate our approach using a large-scale dataset of 500 famous books (predominantly copyrighted works) and demonstrate that SUV significantly reduces verbatim memorization with negligible impact on the performance on unrelated tasks. Extensive experiments on both our dataset and public benchmarks confirm the scalability and efficacy of our approach, offering a promising solution for mitigating copyright risks in real-world LLM applications.

cs.CL↗

arXiv2025

Towards Universal Debiasing for Language Models-based Tabular Data Generation

Large language models (LLMs) have achieved promising results in tabular data generation. However, inherent historical biases in tabular datasets often cause LLMs to exacerbate fairness issues, particularly when multiple advantaged and protected features are involved. In this work, we introduce a universal debiasing framework that minimizes group-level dependencies by simultaneously reducing the mutual information between advantaged and protected attributes. By leveraging the autoregressive structure and analytic sampling distributions of LLM-based tabular data generators, our approach efficiently computes mutual information, reducing the need for cumbersome numerical estimations. Building on this foundation, we propose two complementary methods: a direct preference optimization (DPO)-based strategy, namely UDF-DPO, that integrates seamlessly with existing models, and a targeted debiasing technique, namely UDF-MIX, that achieves debiasing without tuning the parameters of LLMs. Extensive experiments demonstrate that our framework effectively balances fairness and utility, offering a scalable and practical solution for debiasing in high-stakes applications.

cs.LG↗

arXiv2025

Effect of protection zone on the dynamics of a diffusion-advection population-toxicant model

This paper develops and analyzes a diffusion-advection model coupling population dynamics with toxicant transport, incorporating a boundary protection zone. For both upstream and downstream protection zone configurations, we investigate the combined influence of protected zones and key ecological factors on population persistence or extinction. Employing monotone dynamical system theory and eigenvalue analysis, we establish the global dynamics of the population-toxicant coexistence equilibrium. Furthermore, we characterize the parameter dependence governing the stability of the toxicant-only steady state, specifically examining the protected zone length, toxicant effect coefficient on population growth, per-unit contaminant discharge rate, toxicant input rate, diffusion/advection rates, and population natural growth rate. Finally, numerical simulations reveal the complex interplay between the protection zone and toxicant advection rate in significantly shaping population persistence domains.

q-bio.PE↗

arXiv2025

Precision calculations of $B\to K^*$ form factors from SCET sum rules beyond leading-power contributions

We construct light-cone sum rules (LCSR) for the $B\to K^*$ form factors in the large recoil region using vacuum-to-$B$-meson correlation functions, and systematically calculate subleading-power corrections to these form factors at tree level, including next-to-leading power contributions from the hard-collinear propagator, the subleading effective current $\bar{q}Γ[i\slashed{D}_{\perp}/(2m_b)]h_v$, and twist-five/six four-particle higher-twist effects. By incorporating the available leading-power results at $\mathcal{O}(α_s)$ and the corrections to higher-twist $B$-meson light-cone distribution amplitudes from our previous work, we improve the precision of theoretical predictions for $B\to K^*$ form factors and find that the subleading-power contributions amount to 30\% of the corresponding leading-power results. Employing the Bourrely-Caprini-Lellouch (BCL) parametrization, we determine the numerical results for $B\to K^*$ form factors across the full kinematic range through a combined fit of LCSR predictions in the large recoil region and lattice QCD results in the small recoil region. Using the newly obtained $B\to K^*$ form factors, we compute the branching fractions for the rare decays $B \to K^* ν_\ell\barν_\ell$ in the Standard Model, obtaining $\mathcal{BR}(\bar{B}^0 \to \bar{K}^{*0} ν_\ell\barν_\ell)=8.09(96)\times 10^{-6}$ and $\mathcal{BR}(\bar{B}^+ \to \bar{K}^{*+} ν_\ell\barν_\ell)=9.95(1.05)\times 10^{-6}$. Additionally, we predict that the longitudinal $K^*$ polarization fraction is $F_L=0.44(4)$.

hep-ph↗

arXiv2025

Towards AI-Native RAN: An Operator's Perspective of 6G Day 1 Standardization

Artificial Intelligence/Machine Learning (AI/ML) has become the most certain and prominent feature of 6G mobile networks. Unlike 5G, where AI/ML was not natively integrated but rather an add-on feature over existing architecture, 6G shall incorporate AI from the onset to address its complexity and support ubiquitous AI applications. Based on our extensive mobile network operation and standardization experience from 2G to 5G, this paper explores the design and standardization principles of AI-Native radio access networks (RAN) for 6G, with a particular focus on its critical Day 1 architecture, functionalities and capabilities. We investigate the framework of AI-Native RAN and present its three essential capabilities to shed some light on the standardization direction; namely, AI-driven RAN processing/optimization/automation, reliable AI lifecycle management (LCM), and AI-as-a-Service (AIaaS) provisioning. The standardization of AI-Native RAN, in particular the Day 1 features, including an AI-Native 6G RAN architecture, were proposed. For validation, a large-scale field trial with over 5000 5G-A base stations have been built and delivered significant improvements in average air interface latency, root cause identification, and network energy consumption with the proposed architecture and the supporting AI functions. This paper aims to provide a Day 1 framework for 6G AI-Native RAN standardization design, balancing technical innovation with practical deployment.

cs.NI↗

arXiv2025

Unsupervised 3D Braided Hair Reconstruction from a Single-View Image

Reconstructing 3D braided hairstyles from single-view images remains a challenging task due to the intricate interwoven structure and complex topologies of braids. Existing strand-based hair reconstruction methods typically focus on loose hairstyles and often struggle to capture the fine-grained geometry of braided hair. In this paper, we propose a novel unsupervised pipeline for efficiently reconstructing 3D braided hair from single-view RGB images. Leveraging a synthetic braid model inspired by braid theory, our approach effectively captures the complex intertwined structures of braids. Extensive experiments demonstrate that our method outperforms state-of-the-art approaches, providing superior accuracy, realism, and efficiency in reconstructing 3D braided hairstyles, supporting expressive hairstyle modeling in digital humans.

cs.CV↗

arXiv2025

Observation of High-Order Quantum Pancharatnam-Berry Phase with Structured Photons

When a quantum system evolves so that it returns to its initial state, it will acquire a geometric phase acting as a memory of the transformation of a physical system, which has been experimentally measured in a variety of physical systems. In optics, the most prominent example is the Pancharatnam-Berry (PB) phase. Recent technological advances in phase and polarization structure have led to the discovery of high-order PB phases with structured light fields. The study on the high-order PB phase is limited in the context of elementary quantum states of light, especially in the case of photon number states. Here, we experimentally investigate the differences of high-order PB phases between single-photon and N00N states. Our results show that the PB phase, like the dynamic phase, can also be doubled under two-photon states, which can greatly improve the phase sensitivity for greater $N$ in N00N states and high-order structured photons. This may show some implications for quantum precision measurement and quantum state engineering based on geometric phase.

quant-ph↗

arXiv2025

Mitigating Heterogeneous Token Overfitting in LLM Knowledge Editing

Large language models (LLMs) have achieved remarkable performance on various natural language tasks. However, they are trained on static corpora and their knowledge can become outdated quickly in the fast-changing world. This motivates the development of knowledge editing (KE) to update specific knowledge in LLMs without changing unrelated others or compromising their pre-trained capabilities. Previous efforts sought to update a small amount of parameters of a LLM and proved effective for making selective updates. Nonetheless, the edited LLM often exhibits degraded ability to reason about the new knowledge. In this work, we identify a key issue: heterogeneous token overfitting (HTO), where the LLM overfits different tokens in the provided knowledge at varying rates. To tackle this, we propose OVERTONE, a token-level smoothing method that mitigates HTO by adaptively refining the target distribution. Theoretically, OVERTONE offers better parameter updates with negligible computation overhead. It also induces an implicit DPO but does not require preference data pairs. Extensive experiments across four editing methods, two LLMs, and diverse scenarios demonstrate the effectiveness and versatility of our method.

cs.CL↗

arXiv2025

Realization of universal nonadiabatic holonomic quantum gates in trapped ion system

The implementation of holonomic quantum computation is meaningful. We can effectively resist local and collective noise in the process of physical implementation by using the advantage of non-Abelian geometric phase. In this paper, we set out from the simplest and most fundamental Jaynes-Cummings model of ion trap system to implement single-qubit logical operations, and taking advantage of the interaction of two ions with a pair of laser pulses to implement two-qubit logical operations, respectively. Compared with the previous proposal, the whole process of our proposal are in tunable way and the universal holonomic quantum qubit gates can be made robust to systematic error and decay which pushes the gate fidelities in the presence of decoherence and systematic error to well high level. Thus our scheme affords an experimentally feasible and simple way to make realizing the robust nonadiabatic holonomic quantum computation.

quant-ph↗

arXiv2025

Towards Federated RLHF with Aggregated Client Preference for LLMs

Reinforcement learning with human feedback (RLHF) fine-tunes a pretrained large language model (LLM) using user preference data, enabling it to generate content aligned with human preferences. However, due to privacy concerns, users may be reluctant to share sensitive preference data. To address this, we propose utilizing Federated Learning (FL) techniques, allowing large-scale preference collection from diverse real-world users without requiring them to transmit data to a central server. Our federated RLHF methods (i.e., FedBis and FedBiscuit) encode each client's preferences into binary selectors and aggregate them to capture common preferences. In particular, FedBiscuit overcomes key challenges, such as preference heterogeneity and reward hacking, through innovative solutions like grouping clients with similar preferences to reduce heterogeneity and using multiple binary selectors to enhance LLM output quality. To evaluate the performance of the proposed methods, we establish the first federated RLHF benchmark with a heterogeneous human preference dataset. Experimental results show that by integrating the LLM with aggregated client preferences, FedBis and FedBiscuit significantly enhance the professionalism and readability of the generated content.

cs.CL↗

arXiv2025

DiSRT-In-Bed: Diffusion-Based Sim-to-Real Transfer Framework for In-Bed Human Mesh Recovery

In-bed human mesh recovery can be crucial and enabling for several healthcare applications, including sleep pattern monitoring, rehabilitation support, and pressure ulcer prevention. However, it is difficult to collect large real-world visual datasets in this domain, in part due to privacy and expense constraints, which in turn presents significant challenges for training and deploying deep learning models. Existing in-bed human mesh estimation methods often rely heavily on real-world data, limiting their ability to generalize across different in-bed scenarios, such as varying coverings and environmental settings. To address this, we propose a Sim-to-Real Transfer Framework for in-bed human mesh recovery from overhead depth images, which leverages large-scale synthetic data alongside limited or no real-world samples. We introduce a diffusion model that bridges the gap between synthetic data and real data to support generalization in real-world in-bed pose and body inference scenarios. Extensive experiments and ablation studies validate the effectiveness of our framework, demonstrating significant improvements in robustness and adaptability across diverse healthcare scenarios.

cs.CV↗

arXiv2025

Building Machine Learning Challenges for Anomaly Detection in Science

Scientific discoveries are often made by finding a pattern or object that was not predicted by the known rules of science. Oftentimes, these anomalous events or objects that do not conform to the norms are an indication that the rules of science governing the data are incomplete, and something new needs to be present to explain these unexpected outliers. The challenge of finding anomalies can be confounding since it requires codifying a complete knowledge of the known scientific behaviors and then projecting these known behaviors on the data to look for deviations. When utilizing machine learning, this presents a particular challenge since we require that the model not only understands scientific data perfectly but also recognizes when the data is inconsistent and out of the scope of its trained behavior. In this paper, we present three datasets aimed at developing machine learning-based anomaly detection for disparate scientific domains covering astrophysics, genomics, and polar science. We present the different datasets along with a scheme to make machine learning challenges around the three datasets findable, accessible, interoperable, and reusable (FAIR). Furthermore, we present an approach that generalizes to future machine learning challenges, enabling the possibility of large, more compute-intensive challenges that can ultimately lead to scientific discovery.

cs.LG↗

arXiv2025

The maximum spectral radius of $θ_{2,2,3}$-free graphs with given size

A theta graph $θ_{r,p,q}$ is the graph obtained by connecting two distinct vertices with three internally disjoint paths of length $r,p,q$, where $q\geq p\geq r\geq1$ and $p\geq2$. A graph is $θ_{r,p,q}$-free if it does not contain $θ_{r,p,q}$ as a subgraph. The maximum spectral radius of $θ_{1,p,q}$-free graphs with given size has been determined for any $q\geq p\geq2$. Zhai, Lin and Shu [Spectral extrema of graphs with fixed size: cycles and complete bipartite graphs, European J. Combin. 95 (2021) 103322] characterized the extremal graph with the maximum spectral radius of $θ_{2,2,2}$-free graphs having $m$ edges. In this paper, we consider the maximum spectral radius of $θ_{2,2,3}$-free graphs with size $m$ and characterize the extremal graph.

math.CO↗

arXiv2025

Unlocking Efficient, Scalable, and Continual Knowledge Editing with Basis-Level Representation Fine-Tuning

Large language models (LLMs) have achieved remarkable performance on various natural language tasks. However, they are trained on static corpora and their knowledge can become outdated quickly in the fast-changing world. This motivates the development of knowledge editing methods designed to update certain knowledge in LLMs without changing unrelated others. To make selective edits, previous efforts often sought to update a small amount of parameters in some specific layer(s) of a LLM. Nonetheless, in challenging scenarios, they still fall short in making successful edits while preserving knowledge irrelevant to the updates simultaneously, resulting in a notable editing-locality trade-off. In this work, we question if the trade-offs are caused by the fact that parameter-based updates have a global effect, i.e., edited parameters affect all inputs indiscriminately. In light of this, we explore the feasibility of representation fine-tuning, which applied some linear update to a few representations in a learned subspace, for knowledge editing. While being effective to enhance an LLM's general ability as demonstrated in the previous work, we theoretically show that this linear update imposes a tension in editing-locality trade-off. Subsequently, BaFT is proposed to break the linearity. BaFT computes a weight for each basis that spans a dimension of the subspace based on the input representation. This input-dependent weighting mechanism allows BaFT to manage different types of knowledge in an adaptive way, thereby achieving a better editing-locality trade-off. Experiments on three LLMs with five editing benchmarks in diverse scenarios show the superiority of our method.

cs.CL↗

arXiv2025

Generating Millions Of Lean Theorems With Proofs By Exploring State Transition Graphs

Large Language Models (LLMs) have demonstrated significant potential in generating mathematical proofs. However, a persistent challenge is that LLMs occasionally make mistakes, while even a minor mistake can invalidate an entire proof. Proof assistants like Lean offer a great remedy. They are designed for verifying each step of a proof in a formal language, and in recent years researchers have created AI models to generate proofs in their languages. However, the scarcity of large-scale datasets of Lean proofs restrict the performance of such Automated Theorem Proving (ATP) models. We developed LeanNavigator, a novel method for generating a large-scale dataset of Lean theorems and proofs by finding new ways to prove existing Lean theorems. By leveraging an interactive Lean client and an efficient method for proof step generation, LeanNavigator efficiently produces new theorems with corresponding proofs. Applying this approach to Mathlib4, we generated 4.7 million theorems totaling 1 billion tokens, surpassing previous datasets by more than an order of magnitude. Using this extensive dataset, we trained an AI model that outperforms the state-of-the-art ReProver model in theorem-proving tasks. These results confirm our hypothesis and demonstrate the critical role of large datasets in improving the performance of automated theorem provers.

cs.LO↗

arXiv2025

Counterfactual Fairness by Combining Factual and Counterfactual Predictions

In high-stake domains such as healthcare and hiring, the role of machine learning (ML) in decision-making raises significant fairness concerns. This work focuses on Counterfactual Fairness (CF), which posits that an ML model's outcome on any individual should remain unchanged if they had belonged to a different demographic group. Previous works have proposed methods that guarantee CF. Notwithstanding, their effects on the model's predictive performance remains largely unclear. To fill in this gap, we provide a theoretical study on the inherent trade-off between CF and predictive performance in a model-agnostic manner. We first propose a simple but effective method to cast an optimal but potentially unfair predictor into a fair one without losing the optimality. By analyzing its excess risk in order to achieve CF, we quantify this inherent trade-off. Further analysis on our method's performance with access to only incomplete causal knowledge is also conducted. Built upon it, we propose a performant algorithm that can be applied in such scenarios. Experiments on both synthetic and semi-synthetic datasets demonstrate the validity of our analysis and methods.

cs.LG↗

arXiv2024

Spectral radius of graphs of given size with forbidden a fan graph $F_6$

Let $F_k=K_1\vee P_{k-1}$ be the fan graph on $k$ vertices. A graph is said to be $F_k$-free if it does not contain $F_k$ as a subgraph. Yu et al. in [arXiv:2404.03423] conjectured that for $k\geq2$ and $m$ sufficiently large, if $G$ is an $F_{2k+1}$-free or $F_{2k+2}$-free graph, then $λ(G)\leq \frac{k-1+\sqrt{4m-k^2+1}}{2}$ and the equality holds if and only if $G\cong K_k\vee\left(\frac{m}{k}-\frac{k-1}{2}\right)K_1$. Recently, Li et al. in [arXiv:2409.15918] showed that the above conjecture holds for $k\geq 3$. The only left case is for $k=2$, which corresponds to $F_5$ or $F_6$. Since the case of $F_5$ was solved by Yu et al. in [arXiv:2404.03423] and Zhang and Wang in [On the spectral radius of graphs without a gem, Discrete Math. 347 (2024) 114171]. So, one needs only to deal with the case of $F_6$. In this paper, we solve the only left case by determining the maximum spectral radius of $F_6$-free graphs with size $m\geq 88$, and the corresponding extremal graph.

math.CO↗