arXiv subjects

Fan Lin

Publications and source records attributed to Fan Lin.

At least 19 recordsLinked to original sources

arXiv2026

Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA

Macaron-V1 is an open agent-model family for experiential intelligence: learning from experience in real environments and continuing to learn after deployment. It is organized around two system goals. Adaptation is pursued through recursive improvement of versioned model-harness pairs, where experience from one configuration is evaluated under an external contract and used to construct its successor. Collaboration is pursued via the Mixture-of-LoRA (MoL) architecture that freezes a base model, composes specialist LoRA adapters, and selects one LoRA per user turn. The flagship Macaron-V1-Venti (748B) combines a 744B GLM-5.2 base with four LoRAs for chat, agent, coding, and GenUI; the Qwen3.6-35B-based Macaron-V1-Tall (50B) uses the same design for local deployment. This report presents Macaron-V1 as a co-designed system spanning architecture, algorithms, and infrastructure. The MoL architecture supports continual learning through extensible LoRA specialists. The algorithm combines Model-Harness Co-design and recursive self-improvement loop, including the UI4A component-native GenUI harness, a stateful action substrate, versioned Harness Context Protocol contract, and the agentic RL framework MindForge. The supporting infrastructure includes the post-training platform MinT, the long-context RL method LongStraw, and stability techniques for sparse MoE and DSA base models. We evaluate Macaron-V1 on Personal Intelligence, GenUI, and general capability benchmarks against frontier baselines. Our results validate the current system, while compounding gains from continual learning and collective intelligence remain open questions.

cs.LG↗

arXiv2026

RG-Consistent (P)NJL Model: Impact of Thermal Cutoff Modifications on Thermodynamics and Net-Baryon Number Fluctuations

In this paper, we investigate the impact of renormalization group (RG) consistency on the chiral phase transition and thermodynamic properties of QCD matter using the RGNJL and RGPNJL models. By implementing a temperature-dependent thermal cutoff $\Lambda_T = k\Lambda_0$, we ensure that thermodynamic quantities converge toward the Stefan-Boltzmann limit at high temperatures, effectively extending the applicability of these effective theories. Our analysis shows that while the RG-consistency condition ($k \rightarrow \infty$) resolves causality violations in the RGNJL model by binding the speed of sound to the conformal limit, the RGPNJL model exhibits a more complex, non-monotonic sensitivity to the parameter $k$. Furthermore, we demonstrate that the RG-improved PNJL framework significantly enhances the description of net-baryon number fluctuations ($\kappa\sigma^2$) relative to lattice QCD data at vanishing chemical potential, though the intensification of these fluctuations at high baryon density highlights a critical sensitivity to the model's parametric constraints. This study provides a rigorous evaluation of the RG-consistency framework's predictive power in mapping the QCD phase diagram and interpreting experimental observables.

hep-ph↗

arXiv2026

Federated Nested Learning: Collaborative Training of Self-Referential Memories for Test-Time Adaptation

We rethink Federated Learning (FL) from a nested learning perspective, framing the core challenge as how to collaboratively learn optimization rules, not just static models, to tackle Non-IID client data. To address this, we propose Federated Nested Learning (FedNL), a novel framework that reformulates FL as a three-level nested optimization system. FedNL embeds Titans-based linear attention into FL, enabling clients to perform lightweight, zero-shot test-time adaptation by treating a delta rule as an online gradient step. Experiments on Non-IID MMLU and long-context benchmarks show that FedNL achieves competitive performance in short-context reasoning, enhances the performance of long-context retrieval and streaming Cross-Entropy, and maintains constant inference memory.

cs.LG↗

arXiv2026

Determining the NJL Coupling and AMM in Magnetized QCD Matter via Machine Learning

In this study, we investigate the phase structure of magnetized QCD matter by determining the field-dependent parameters of the Nambu-Jona-Lasinio (NJL) model through a physics-informed machine learning framework. Specifically, we focus on extracting the optimal functional forms for the running coupling constant $G(eB)$ and the quark anomalous magnetic moment (AMM) ratio $v_2(eB)$, utilizing lattice QCD-computed quark condensate data as the ``ground truth". By embedding the NJL gap equation as a differentiable physics-constrained module, our neural network pipeline identifies continuous parameter functions that accurately reproduce the inverse magnetic catalysis (IMC) effect. Our results demonstrate that the magnetic field smoothly suppresses both $G$ and $v_2$. This approach not only bridges the gap between effective models and lattice data but also provides new microscopic insights into the response of the QCD vacuum to strong magnetic fields.

hep-ph↗

arXiv2026

Toward a Unified Picture of Confinement and Baryon Structure

In this work, we investigate the infrared structure of quantum chromodynamics from the perspective of the Cho-Faddeev-Niemi decomposition and the Faddeev-Niemi effective theory of Yang-Mills fields. We argue that the topological solitons of the Faddeev-Niemi theory, namely gluon knots characterized by the Hopf invariant, should be regarded as the relevant ground-state degrees of freedom of Yang-Mills theory in the deep infrared region. In this framework, gluon knots provide a unified description of monopole condensation and the center-vortex confinement mechanism. We further propose that baryons are composite objects consisting of quarks immersed in a gluon-knot background. The monopole condensate associated with the gluon knot realizes dual superconductivity, squeezes color-electric flux into flux tubes, and naturally generates the $\mathrm{Y}$-shaped confinement structure of baryons. Simultaneously, the strong local color-magnetic field generated by the gluon knot induces chiral symmetry breaking through magnetic catalysis and topological vacuum fluctuations. We show that the Hopf invariant of the gluon knot is closely related to the topology of Yang-Mills vacua and discuss its connection with the axial anomaly and instanton-induced chiral symmetry breaking. Furthermore, the topological current associated with gluon knots provides a natural carrier of gluon angular momentum and may account for a substantial fraction of the proton spin. The resulting picture establishes a possible connection between the infrared topology of Yang-Mills theory and the internal structure of baryons, providing a unified framework for confinement, chiral symmetry breaking, and baryon structure in QCD.

hep-th↗

arXiv2025

Photon emission from weakly magnetized neutral pions

Using a hadronic framework, we derive an explicit expression for photon production from neutral pions in a weak background magnetic field. Our calculation is built on the proton triangle diagram with an effective Yukawa $\pi^0$-proton coupling, offering an alternative to quark-level descriptions that is advantageous when the magnetic length greatly exceeds the proton size. Corrections to the pion decay constant are computed up to second order in the magnetic-field strength, revealing that the field generally suppresses the decay rate. Quantitatively, however, the effect remains modest even for fields as strong as $|eB|\simeq m_\pi^2$. The differential photon emission rate exhibits anisotropy, with the strongest suppression occurring when the pion momentum is perpendicular to the magnetic field. Overall, the modification of the $\pi^0 \to \gamma\gamma$ rate is parametrically small, scaling as $|eB|^2/m_P^4$, where $m_P$ is the proton mass. While the magnetic-field-induced anisotropy is conceptually interesting in principle, it is likely too small to be resolved in present heavy-ion measurements.

hep-ph↗

arXiv2025

A Survey on Data Security in Large Language Models

Large Language Models (LLMs), now a foundation in advancing natural language processing, power applications such as text generation, machine translation, and conversational systems. Despite their transformative potential, these models inherently rely on massive amounts of training data, often collected from diverse and uncurated sources, which exposes them to serious data security risks. Harmful or malicious data can compromise model behavior, leading to issues such as toxic output, hallucinations, and vulnerabilities to threats such as prompt injection or data poisoning. As LLMs continue to be integrated into critical real-world systems, understanding and addressing these data-centric security risks is imperative to safeguard user trust and system reliability. This survey offers a comprehensive overview of the main data security risks facing LLMs and reviews current defense strategies, including adversarial training, RLHF, and data augmentation. Additionally, we categorize and analyze relevant datasets used for assessing robustness and security across different domains, providing guidance for future research. Finally, we highlight key research directions that focus on secure model updates, explainability-driven defenses, and effective governance frameworks, aiming to promote the safe and responsible development of LLM technology. This work aims to inform researchers, practitioners, and policymakers, driving progress toward data security in LLMs.

cs.CR↗

arXiv2025

Diagnosing Failures in Large Language Models' Answers: Integrating Error Attribution into Evaluation Framework

With the widespread application of Large Language Models (LLMs) in various tasks, the mainstream LLM platforms generate massive user-model interactions daily. In order to efficiently analyze the performance of models and diagnose failures in their answers, it is essential to develop an automated framework to systematically categorize and attribute errors. However, existing evaluation models lack error attribution capability. In this work, we establish a comprehensive Misattribution Framework with 6 primary and 15 secondary categories to facilitate in-depth analysis. Based on this framework, we present AttriData, a dataset specifically designed for error attribution, encompassing misattribution, along with the corresponding scores and feedback. We also propose MisAttributionLLM, a fine-tuned model on AttriData, which is the first general-purpose judge model capable of simultaneously generating score, misattribution, and feedback. Extensive experiments and analyses are conducted to confirm the effectiveness and robustness of our proposed method.

cs.CL↗

arXiv2025

RemoteChess: Enhancing Older Adults' Social Connectedness via Designing a Virtual Reality Chinese Chess (Xiangqi) Community

The decline of social connectedness caused by distance and physical limitations severely affects older adults' well-being and mental health. While virtual reality (VR) is promising for older adults to socialize remotely, existing social VR designs primarily focus on verbal communication (e.g., reminiscent, chat). Actively engaging in shared activities is also an important aspect of social connection. We designed RemoteChess, which constructs a social community and a culturally relevant activity (i.e., Chinese chess) for older adults to play while engaging in social interaction. We conducted a user study with groups of older adults interacting with each other through RemoteChess. Our findings indicate that RemoteChess enhanced participants' social connectedness by offering familiar environments, culturally relevant social catalysts, and asymmetric interactions. We further discussed design guidelines for designing culturally relevant social activities in VR to promote social connectedness for older adults.

cs.HC↗

arXiv2025

Baryon Construction with $\eta^\prime$ Meson Field

In the low-energy regime, baryons with $N_f \geq 2$ have long been constructed as skyrmions or through bag models, but such constructions for $N_f = 1$ are hindered by the trivial topological structure of the meson field. Recent proposals suggest that one-flavor baryons can instead be interpreted as quantum Hall droplets on the $\eta'$ domain wall, providing a potential link to quark--hadron continuity at high density. In retrospect, the qualitative or semi-qualitative construction of one-flavor baryons on the $\eta'$ domain wall reveals that these baryons can be described as quantum Hall droplets, resembling topological solitons akin to skyrmions. Using an effective theory on the $\eta'$ domain wall, which is conjectured to be the Chern--Simons--Higgs theory, it is discussed that its vortex solution with unit baryon numbers naturally has a spin of $N_c/2$, and thus can be interpreted as a baryon or multi-baryon structure. The particle--vortex duality suggests that quarks carry a fractional topological charge of $1/N_c$ and obey fractional statistics. In terms of chiral bag models, confinement can be attributed to the monopoles confined within the bag, and the vector meson fields on the bag surface are essential for ensuring the correct baryon number in the chiral bag framework, thereby providing deeper insights into baryons as non-trivial topological structures of the meson field. In this paper, we review the progress in this development, with a special focus on the $\eta^\prime$ domain wall dynamics. Naive extensions to $N_f \geq 2$ are also discussed.

hep-ph↗

arXiv2025

Confined Monopoles in Chiral Bag

The chiral bag model offers a dual description of hadron physics in terms of quarks and hadrons in the sense of Cheshire Cat principle. In this work, we find that, within the chiral bag, confinement is likely caused by monopole condensation. The chiral bag surface can be interpreted as an $\eta'$ domain wall, where a dynamical Chern-Simons theory emerges. Under level-rank duality, the Chern-Simons theory serves as the counterterm introduced to block the so-called color charge leakage. To ensure the correct net baryon number of the full chiral bag, an additional Chern-Simons theory involving the vector meson field arises on the bag surface and extends outside the bag. This leads to a Chern-Simons-Higgs theory localized on the $\eta'$ domain wall, as previously conjectured. We also propose that the skyrmion description of baryons could be understood as a system of monopoles enveloped by a meson cloud.

hep-ph↗

arXiv2024

IDGen: Item Discrimination Induced Prompt Generation for LLM Evaluation

As Large Language Models (LLMs) grow increasingly adept at managing complex tasks, the evaluation set must keep pace with these advancements to ensure it remains sufficiently discriminative. Item Discrimination (ID) theory, which is widely used in educational assessment, measures the ability of individual test items to differentiate between high and low performers. Inspired by this theory, we propose an ID-induced prompt synthesis framework for evaluating LLMs to ensure the evaluation set can continually update and refine according to model abilities. Our data synthesis framework prioritizes both breadth and specificity. It can generate prompts that comprehensively evaluate the capabilities of LLMs while revealing meaningful performance differences between models, allowing for effective discrimination of their relative strengths and weaknesses across various tasks and domains. To produce high-quality data, we incorporate a self-correct mechanism into our generalization framework, and develop two models to predict prompt discrimination and difficulty score to facilitate our data synthesis framework, contributing valuable tools to evaluation data synthesis research. We apply our generated data to evaluate five SOTA models. Our data achieves an average score of 51.92, accompanied by a variance of 10.06. By contrast, previous works (i.e., SELF-INSTRUCT and WizardLM) obtain an average score exceeding 67, with a variance below 3.2. The results demonstrate that the data generated by our framework is more challenging and discriminative compared to previous works. We will release a dataset of over 3,000 carefully crafted prompts to facilitate evaluation research of LLMs.

cs.CL↗

arXiv2024

Evolutionary Reinforcement Learning: A Systematic Review and Future Directions

In response to the limitations of reinforcement learning and evolutionary algorithms (EAs) in complex problem-solving, Evolutionary Reinforcement Learning (EvoRL) has emerged as a synergistic solution. EvoRL integrates EAs and reinforcement learning, presenting a promising avenue for training intelligent agents. This systematic review firstly navigates through the technological background of EvoRL, examining the symbiotic relationship between EAs and reinforcement learning algorithms. We then delve into the challenges faced by both EAs and reinforcement learning, exploring their interplay and impact on the efficacy of EvoRL. Furthermore, the review underscores the need for addressing open issues related to scalability, adaptability, sample efficiency, adversarial robustness, ethic and fairness within the current landscape of EvoRL. Finally, we propose future directions for EvoRL, emphasizing research avenues that strive to enhance self-adaptation and self-improvement, generalization, interpretability, explainability, and so on. Serving as a comprehensive resource for researchers and practitioners, this systematic review provides insights into the current state of EvoRL and offers a guide for advancing its capabilities in the ever-evolving landscape of artificial intelligence.

cs.NE↗

arXiv2023

Baryons as Vortexes on the $\eta^{\prime}$ Domain Wall

We show that the recent construction of $N_f=1$ baryons on the $\eta^\prime$ domain wall can be understood as vortexes of the principal effective theory -- the Chern-Simons-Higgs theory -- on a 2+1-dimensional sheet. This theory has a series of vertex solutions, and the vortex with unit topological charge naturally spins $N_c/2$, which coincides with the spin of the one-flavor baryon in QCD. Since the $N_c$ scaling of the vortexes is the same as that of baryons, baryons can be regarded as vortexes. By virtue of the particle-vortex symmetry, the dual Zhang-Hansson-Kivelson theory indicates that the quark carries topological charge $1/N_c$ and obeys fractional statistics. The generalization to arbitrary $N_f$ is also discussed.

hep-th↗

arXiv2023

A Comprehensive Survey on Deep Learning Techniques in Educational Data Mining

Educational Data Mining (EDM) has emerged as a vital field of research, which harnesses the power of computational techniques to analyze educational data. With the increasing complexity and diversity of educational data, Deep Learning techniques have shown significant advantages in addressing the challenges associated with analyzing and modeling this data. This survey aims to systematically review the state-of-the-art in EDM with Deep Learning. We begin by providing a brief introduction to EDM and Deep Learning, highlighting their relevance in the context of modern education. Next, we present a detailed review of Deep Learning techniques applied in four typical educational scenarios, including knowledge tracing, student behavior detection, performance prediction, and personalized recommendation. Furthermore, a comprehensive overview of public datasets and processing tools for EDM is provided. We then analyze the practical challenges in EDM and propose targeted solutions. Finally, we point out emerging trends and future directions in this research area.

cs.LG↗

arXiv2023

Diffusion Model for Camouflaged Object Detection

Camouflaged object detection is a challenging task that aims to identify objects that are highly similar to their background. Due to the powerful noise-to-image denoising capability of denoising diffusion models, in this paper, we propose a diffusion-based framework for camouflaged object detection, termed diffCOD, a new framework that considers the camouflaged object segmentation task as a denoising diffusion process from noisy masks to object masks. Specifically, the object mask diffuses from the ground-truth masks to a random distribution, and the designed model learns to reverse this noising process. To strengthen the denoising learning, the input image prior is encoded and integrated into the denoising diffusion model to guide the diffusion process. Furthermore, we design an injection attention module (IAM) to interact conditional semantic features extracted from the image with the diffusion noise embedding via the cross-attention mechanism to enhance denoising learning. Extensive experiments on four widely used COD benchmark datasets demonstrate that the proposed method achieves favorable performance compared to the existing 11 state-of-the-art methods, especially in the detailed texture segmentation of camouflaged objects. Our code will be made publicly available at: https://github.com/ZNan-Chen/diffCOD.

cs.CV↗

arXiv2022

$\Lambda/{\bar \Lambda}$ Polarization and Splitting Induced by Rotation and Magnetic Field

The global polarization of $\Lambda/{\bar \Lambda}$ and the splitting of ${\bar \Lambda}-\Lambda$ polarization induced by rotation and magnetic field has been investigated in a dynamical quark model by taking into account the axial vector interaction and the anomalous magnetic moment of quarks. It is found that the rotation leads to the spin polarization of quarks and anti-quarks with the same sign, while the magnetic field to opposite sign, which corresponds to the ${\bar \Lambda}-\Lambda$ polarization splitting. The combination of the two effects leads to perfect agreement with experiment data. Quantitatively, the axial vector spin polarization contributes 30$\%$ of the global polarization and the anomalous magnetic moment of quarks contributes 40$\%$ to the splitting of ${\bar \Lambda}-\Lambda$ polarization. However, at $\sqrt{s_{NN}} \leq 7.7 \text{GeV}$, it still remains a challenge to reach enough magnitude of the magnetic field at freeze-out.

hep-ph↗

arXiv2022

Magnetism of QCD matter and pion mass from tensor-type spin polarization and anomalous magnetic moment of quarks

We investigate the magnetism of QCD matter and pion mass under magnetic field considering the contribution from the tensor-type spin polarization and the anomalous magnetic moment (AMM) of quarks. It is found that the tensor-type spin polarization (TSP) induces the magnetic catalysis of chiral condensate and diamagnetism (negative magnetic susceptibility) of quark matter at low temperature, both neutral and charged pion masses increase quickly with magnetic field in the case of TSP. The anomalous magnetic moment (AMM) of quarks induces magnetic inhibition and a magnetic dependent AMM causes inverse magnetic catalysis at finite temperature, and the neutral pion mass decreases with magnetic field while the charged pion mass shows nonmonotonic behavior with the magnetic field, which is qualitatively in agreement with lattice result. However, the magnetic susceptibility is positive at low temperature with AMM. In the current framework, our results show the irreconcilable contradiction between the diamagnetism and inverse magnetic catalysis.

hep-ph↗