arXiv subjects

Junyi Cheng

Publications and source records attributed to Junyi Cheng.

16 recordsLinked to original sources

arXiv2026

From Empathy to Personalized Empathy: Adapting Empathetic Strategies to Individual Users

As Large Language Models (LLMs) are increasingly deployed in long-term interactions with users, empathy has become an increasingly important capability. However, existing research overlooks the influence of users' personality traits on empathetic strategies during long-term interactions. To address this gap, we introduce the task of personalized empathy, which focuses on adapting empathetic strategies according to users' personalized characteristics derived from history. To study and enhance this capability, we construct PersonaEmp, a personalized empathy dataset built from long-term user-AI interactions, featuring rich user histories, persona information, and empathy-seeking queries. We further propose PereGRM, a reward modeling framework that combines the empathy evaluation structure with dynamic evaluation criteria generation for fine-grained reward modeling. Experimental results across different settings and multiple judge models show that PereGRM consistently achieves the strongest performance improvements, indicating its effectiveness for enhancing personalized empathetic capabilities.

cs.CL↗

arXiv2026

A coherent structure transport model for scrape-off layer turbulence

Understanding the locality of high-temperature plasma energy deposition on material surfaces in fusion reactors is critical for design. Here, we utilize the Gyrokinetic ElectroMagnetic turbulence including X-points (GEMX) simulation, together with SOLPS-ITER solutions for the background equilibrium electric field including drifts, to model the heat flux at the divertor plate and characterize the heat load width using realistic X-point geometry. We use a theory-based blobby transport model called the "Coherent Structure Transport" (CST) model to include the effect of plasma transport in the edge scrape-off layer. The CST model is extremely fast and can be used to quickly analyze any SOLPS-ITER solution. SOLPS-ITER provides the steady state, or equilibrium on which we superimpose blobby turbulence characterized by blob size, amplitude and frequency. We obtain the $1/B_p$ scaling of the heat load exponential decay width $\lambda_q$, in agreement with the Eich empirical scaling and with the Goldston heuristic theory. When including blobby turbulence in combination with the SOLPS-ITER electric field, we find a secondary peak in the heat flux radial profile, outwardly displaced from the strike point radius, with a relative amplitude that increases with the initial blob density. We describe the CST model in detail and provide initial investigations of the scaling of $\lambda_q$ and the secondary heat flux peak with blob size and amplitude.

physics.plasm-ph↗

arXiv2026

PERM: Psychology-grounded Empathetic Reward Modeling for Large Language Models

Large Language Models (LLMs) are increasingly deployed in human-centric applications, yet they often fail to provide substantive emotional support. While Reinforcement Learning (RL) has been utilized to enhance empathy of LLMs, existing reward models typically evaluate empathy from a single perspective, overlooking the inherently bidirectional interaction nature of empathy between the supporter and seeker as defined by Empathy Cycle theory. To address this limitation, we propose Psychology-grounded Empathetic Reward Modeling (PERM). PERM operationalizes empathy evaluation through a bidirectional decomposition: 1) Supporter perspective, assessing internal resonation and communicative expression; 2) Seeker perspective, evaluating emotional reception. Additionally, it incorporates a bystander perspective to monitor overall interaction quality. Extensive experiments on a widely-used emotional intelligence benchmark and an industrial daily conversation dataset demonstrate that PERM outperforms state-of-the-art baselines by over 10\%. Furthermore, a blinded user study reveals a 70\% preference for our approach, highlighting its efficacy in generating more empathetic responses. Our code, dataset, and models are available at https://github.com/ZhengWwwq/PERM.

cs.CL↗

arXiv2025

Universality in the Axiverse

Studies of axion effective theories in the type IIB Calabi-Yau landscape have revealed that hierarchies in cycle volumes drive correlations between axion physics and the number of axions in a given model. We analyze distributions of divisor volumes in toric hypersurface Calabi-Yau threefolds, and provide evidence that aspects of these distributions are universal across this landscape. Furthermore, we show that axion observables in this landscape can be approximated efficiently in terms of only divisor volumes. Finally, we propose a simple model for the spectrum of divisor volumes and use this model to reproduce results on axion masses and decay constants across the Kreuzer-Skarke axiverse.

hep-th↗

arXiv2025

A Step in Flux to Suppress Axion Isocurvature

The QCD axion in the pre-inflation scenario faces a stringent isocurvature constraint, which requires a relatively low Hubble scale during inflation. If the axion was heavier than the Hubble scale during inflation, its isocurvature is suppressed and the constraint disappears. We point out a novel mechanism for achieving this, relying on the topological nature of a BF-type (monodromy) mass for the axion. Such a mass term has an integer coefficient, so it could naturally have been very large during inflation and exactly zero by the time of the QCD phase transition. This integer can be viewed as a quantized flux, which is discharged in a first-order phase transition that proceeds by the nucleation of charged branes. This mechanism can be embedded in cosmology in several different ways, with tunneling during, at the end of, or after inflation. We provide a detailed case study of the scenario in which the tunneling event occurs during inflation. We also comment briefly on possible UV completions within extra-dimensional gauge theories and string theory. Intriguingly, the phase transition could be accompanied by the emergence of the chiral Standard Model field content from a non-chiral theory during inflation.

hep-ph↗

arXiv2024

Limits on an Exotic Higgs Decay From a Recast ATLAS Four-Lepton Analysis

The ATLAS collaboration, using 139 fb$^{-1}$ of 13 TeV collisions from the Large Hadron Collider, has placed limits on the decay of a $Z$ boson to three dark photons. We reproduce the results of the ATLAS analysis, and then recast it as a limit on a exotic Higgs decay mode, in which the Higgs boson decays via a pair of intermediate (pseudo)scalars $a$ to four dark photons $V$ (or some other spin-one meson). Across the mass range for $m_a$ and $m_V$, we find limits on the exotic Higgs branching fraction BR$(H\to aa \to VVVV)$ in the range of $4\times 10^{-5}$ to $1 \times 10^{-4}$.

hep-ph↗

arXiv2024

The Phase Space Distance Between Collider Events

How can one fully harness the power of physics encoded in relativistic $N$-body phase space? Topologically, phase space is isomorphic to the product space of a simplex and a hypersphere and can be equipped with explicit coordinates and a Riemannian metric. This natural structure that scaffolds the space on which all collider physics events live opens up new directions for machine learning applications and implementation. Here we present a detailed construction of the phase space manifold and its differential line element, identifying particle ordering prescriptions that ensure that the metric satisfies necessary properties. We apply the phase space metric to several binary classification tasks, including discrimination of high-multiplicity resonance decays or boosted hadronic decays of electroweak bosons from QCD processes, and demonstrate powerful performance on simulated data. Our work demonstrates the many benefits of promoting phase space from merely a background on which calculations take place to being geometrically entwined with a theory's dynamics.

hep-ph↗

arXiv2023

Global simulations of kinetic-magnetohydrodynamic processes with energetic electrons in tokamak plasmas

The energetic electrons (EEs) generated through auxiliary heating have been found to destabilize various Alfven eigenmodes (AEs) in recent experiments, which in turn lead to the EE transport and degrade the plasma energy confinement. In this work, we propose a global fluid-kinetic hybrid model for studying corresponding kinetic-magnetohydrodynamic (MHD) processes by coupling the drift-kinetic EEs to the Landau-fluid model of bulk plasmas in a non-perturbative manner. The numerical capability of Landau-fluid bulk plasmas is obtained based on a well-benchmarked eigenvalue code MAS [Multiscale Analysis of plasma Stabilities, J. Bao et al. Nucl. Fusion accepted 2023], and the EE responses to the electromagnetic fluctuations are analytically derived, which not only contribute to the MHD interchange drive and parallel current but also lead to the newly kinetic particle compression with the precessional drift resonance in the leading order. The hybrid model is casted into a nonlinear eigenvalue matrix equation and solved iteratively using Newton's method. By calibrating the EE precession frequency against the particle equation of motion in general geometry and applying more realistic trapped particle distribution in the poloidal plane, MAS simulations of EE-driven beta-induced Alfven eigenmodes (e-BAE) show excellent agreements with gyrokinetic particle-in-cell simulations, and the non-perturbative effects of EEs on e-BAE mode structure, growth rate and damping rate are demonstrated. With these efforts, the upgraded MAS greatly improves the computation efficiency for plasma problems related to deeply-trapped EEs, which is superior than initial-value simulations restricted by the stringent electron Courant condition regarding to the practical application of fast linear analysis.

physics.plasm-ph↗

arXiv2023

MAS: A versatile Landau-fluid eigenvalue code for plasma stability analysis in general geometry

We have developed a new global eigenvalue code, Multiscale Analysis for plasma Stabilities (MAS), for studying plasma problems with wave toroidal mode number n and frequency omega in a broad range of interest in general tokamak geometry, based on a five-field Landau-fluid description of thermal plasmas. Beyond keeping the necessary plasma fluid response, we further retain the important kinetic effects including diamagnetic drift, ion finite Larmor radius, finite parallel electric field, ion and electron Landau resonances in a self-consistent and non-perturbative manner without sacrificing the attractive efficiency in computation. The physical capabilities of the code are evaluated and examined in the aspects of both theory and simulation. In theory, the comprehensive Landau-fluid model implemented in MAS can be reduced to the well-known ideal MHD model, electrostatic ion-fluid model, and drift-kinetic model in various limits, which clearly delineates the physics validity regime. In simulation, MAS has been well benchmarked with theory and other gyrokinetic and kinetic-MHD hybrid codes in a manner of adopting the unified physical and numerical framework, which covers the kinetic Alfven wave, ion sound wave, low-n kink, high-n ion temperature gradient mode and kinetic ballooning mode. Moreover, MAS is successfully applied to model the Alfven eigenmode (AE) activities in DIII-D discharge #159243, which faithfully captures the frequency sweeping of RSAE, the tunneling damping of TAE, as well as the polarization characteristics of KBAE and BAAE being consistent with former gyrokinetic theory and simulation. With respect to the key progress contributed to the community, MAS has the advantage of combining rich physics ingredients, realistic global geometry and high computation efficiency together for plasma stability analysis in linear regime.

physics.plasm-ph↗

arXiv2023

Transport Barrier and Spinning Blob Dynamics in the Tokamak Edge

In this work, we investigate the dynamics of plasma blobs in the edge of magnetic confinement devices using a full-f gyrokinetic particle-in-cell code with X-point geometry. In simulations, the evolution of a seeded blob is followed as it approaches a naturally-forming zonal shear layer near the separatrix, where the blob is stabilized by a large spin induced by the self-consistent adiabatic electron response, and blob bifurcation and trapping are observed during the cross-field propagation of blobs. A new theoretical explanation in both the zonal free and zonal shear layer is constructed, where the dominant ExB spin motion is included. A theoretical condition for a transport barrier induced by the interaction between spinning blobs and the zonal shear layer is obtained, and its scaling is verified with simulations. The new theoretical framework, especially the transport barrier, is applicable to explain and predict various experimental phenomena. In particular, the transport barrier condition calculated with experimental parameters demonstrates that the blob radial transport for H mode is smaller than L mode in experiments.

physics.plasm-ph↗

arXiv2022

An unsupervised approach for semantic place annotation of trajectories based on the prior probability

Semantic place annotation can provide individual semantics, which can be of great help in the field of trajectory data mining. Most existing methods rely on annotated or external data and require retraining following a change of region, thus preventing their large-scale applications. Herein, we propose an unsupervised method denoted as UPAPP for the semantic place annotation of trajectories using spatiotemporal information. The Bayesian Criterion is specifically employed to decompose the spatiotemporal probability of the candidate place into spatial probability, duration probability, and visiting time probability. Spatial information in ROI and POI data is subsequently adopted to calculate the spatial probability. In terms of the temporal probabilities, the Term Frequency Inverse Document Frequency weighting algorithm is used to count the potential visits to different place types in the trajectories, and generates the prior probabilities of the visiting time and duration. The spatiotemporal probability of the candidate place is then combined with the importance of the place category to annotate the visited places. Validation with a trajectory dataset collected by 709 volunteers in Beijing showed that our method achieved an overall and average accuracy of 0.712 and 0.720, respectively, indicating that the visited places can be annotated accurately without any external data.

cs.LG↗

arXiv2021

A New Paradigm for Fast and Repetitive Chirping of Alfv\'en Eigenmodes

A novel 2D nonlinear dynamical paradigm is constructed to interpret the fast and repetitive frequency chirping and amplitude oscillation of Alfv\'en eigenmodes excited by energetic particles in fusion plasmas as observed in global gyrokinetic simulations. In this non-perturbative paradigm of the collisionless phase-space dynamics, wave-particle resonant interactions cause the phase-space structure to continuously twist and fold, leading to the repetitive excitation and decay of the Alfv\'en eigenmode. The radial (perpendicular to the dominant wave-particle interaction) dependence of the mode amplitude and toroidal precessional drifts of the energetic particles leads to the 2D dynamics of wave-particle interactions, which is found to be responsible for the repetitive process of formation and destruction of the mode structure.

physics.plasm-ph↗

arXiv2021

Which Metric on the Space of Collider Events?

Which is the best metric for the space of collider events? Motivated by the success of the Energy Mover's Distance in characterizing collider events, we explore the larger space of unbalanced optimal transport distances, of which the Energy Mover's Distance is a particular case. Geometric and computational considerations favor an unbalanced optimal transport distance known as the Hellinger-Kantorovich distance, which possesses a Riemannian structure that lends itself to efficient linearization. We develop the particle linearized unbalanced Optimal Transport (pluOT) framework for collider events based on the linearized Hellinger-Kantorovich distance and demonstrate its efficacy in boosted jet tagging. This provides a flexible and computationally efficient optimal transport framework ideally suited for collider physics applications.

hep-ph↗

arXiv2021

The Muon Smasher's Guide

We lay out a comprehensive physics case for a future high-energy muon collider, exploring a range of collision energies (from 1 to 100 TeV) and luminosities. We highlight the advantages of such a collider over proposed alternatives. We show how one can leverage both the point-like nature of the muons themselves as well as the cloud of electroweak radiation that surrounds the beam to blur the dichotomy between energy and precision in the search for new physics. The physics case is buttressed by a range of studies with applications to electroweak symmetry breaking, dark matter, and the naturalness of the weak scale. Furthermore, we make sharp connections with complementary experiments that are probing new physics effects using electric dipole moments, flavor violation, and gravitational waves. An extensive appendix provides cross section predictions as a function of the center-of-mass energy for many canonical simplified models.

hep-ph↗

arXiv2021

The Linearized Hellinger--Kantorovich Distance

In this paper we study the local linearization of the Hellinger--Kantorovich distance via its Riemannian structure. We give explicit expressions for the logarithmic and exponential map and identify a suitable notion of a Riemannian inner product. Samples can thus be represented as vectors in the tangent space of a suitable reference measure where the norm locally approximates the original metric. Working with the local linearization and the corresponding embeddings allows for the advantages of the Euclidean setting, such as faster computations and a plethora of data analysis tools, whilst still enjoying approximately the descriptive power of the Hellinger--Kantorovich metric.

math.OC↗

arXiv2020

Linearized Optimal Transport for Collider Events

We introduce an efficient framework for computing the distance between collider events using the tools of Linearized Optimal Transport (LOT). This preserves many of the advantages of the recently-introduced Energy Mover's Distance, which quantifies the "work" required to rearrange one event into another, while significantly reducing the computational cost. It also furnishes a Euclidean embedding amenable to simple machine learning algorithms and visualization techniques, which we demonstrate in a variety of jet tagging examples. The LOT approximation lowers the threshold for diverse applications of the theory of optimal transport to collider physics.

hep-ph↗