Searcharxiv⌕ Search

arXiv subjects

Kun Xu

Publications and source records attributed to Kun Xu.

At least 19 recordsLinked to original sources

Kinetic Non-Equilibrium Receptivity Analysis of Hypersonic Flows via a Complex Linearized Unified Gas-Kinetic Scheme

In the near-space regime, local rarefaction near the leading-edge shock and wall degrades Navier-Stokes-based stability and receptivity analyses due to velocity slip, temperature jumps, and non-equilibrium transport. The unified gas-kinetic scheme (UGKS) provides a multiscale kinetic description spanning continuum to rarefied regimes. Building on this, a complex linearized UGKS (CLUGKS) is developed: first-order perturbations of the discrete UGKS residual are formulated under a time-harmonic assumption with complex amplitudes, converting disturbance equations into a steady elliptic system solved via pseudo-time marching with LU-SGS acceleration. Consequently, kinetic shock structures, wall slip, temperature jumps, and non-Fourier heat fluxes are inherently captured without hybrid coupling. Three benchmarks verify the framework: a standing acoustic wave at an isothermal wall, a Mach 2 normal shock interacting with an obliquely incident acoustic wave, and freestream acoustic receptivity of a hypersonic blunt wedge. The scheme is applied to argon blunt-wedge flows at equivalent altitudes of 40 km and 50 km. At 40 km, UGKS and Navier-Stokes wall responses agree well in period and phase, while wall slip sustains a weakly positive spatial growth rate in UGKS downstream. At 50 km, intensified rarefaction produces a kinetic temperature precursor and a thickened, merged shock layer; while wall-pressure responses remain close, significant deviations emerge in temperature and heat-flux disturbances. These findings show that while continuum models capture dominant wall-pressure signatures under moderate rarefaction, resolving kinetic non-equilibrium across the shock-particularly its thermal footprint-is essential at higher altitudes.

physics.flu-dyn↗

Interplay of Scalar and Pseudo-Scalar Mesonic Channels in $Λ\barΛ$ Spin Correlations

The spin correlation of $Λ\barΛ$ hyperon pairs produced in heavy-ion collisions is investigated through scalar and pseudo-scalar mesonic channels that mediate the creation of $s\bar{s}$ pairs. Because a $J=0$ meson forces $L=S$ while parity requires $(-1)^{L+1}=+/-1$, the two vertices project the pair onto distinct total-spin states: a scalar ($0^+$) vertex selects the $L=S=1$ configuration with a spin weight of $+1$, whereas a pseudo-scalar ($0^-$) vertex selects the $L=S=0$ configuration with a weight of $-3$; these weights are the traces of the correlation tensor, corresponding to $P_{s\bar{s}}=+1/3$ and $-1$, respectively. Implementing this framework within the three-flavor Nambu-Jona-Lasinio model at the Random-Phase-Approximation level, we compute the scalar and pseudo-scalar spectral weights. Our results show that the pseudo-scalar channel dominates, resulting in a negative net correlation fraction throughout. And the spin correlation also depends on the opening angle: strongest spin correlation at around $θ=0$ while weakest at $θ=π$. Furthermore, we show that the collision-energy dependence of the momentum-integrated spin correlation along the freeze-out line is sensitive to the trajectory's path relative to the chiral phase boundary. Trajectories passing near the crossover region exhibit a non-monotonic structure, providing a sensitive probe for the location of the chiral phase transition.

hep-ph↗

Bridging the Morphology Gap: Adapting VLA Models to Dexterous Manipulation via Intent-Conditioned Fine-Tuning

Pretrained vision-language-action (VLA) models offer semantic and spatial priors for robot manipulation, but their parallel-gripper interfaces leave multi-finger contact coordination implicit. Adapting these models requires both timely grasp transitions and embodiment-specific hand trajectories. We introduce InDex, an intent-conditioned framework that connects these two requirements. A normalized aperture signal, derived from retargeted demonstrations, supervises a scalar grasp intent while preserving full hand-action supervision. Stage 1 learns synchronized end-effector-intent chunks; Stage 2 conditions a diffusion hand decoder on these predictions, visual-language context, and proprioception. Sequential optimization establishes the upstream interface before learning dexterous control with low-rate co-adaptation. Across four simulated tasks, InDex with $π_{0.5}$ achieves 85.8% mean task success, improving the native policy by 35.5 percentage points, and reaches 69.0% on held-out objects with similar geometry without fine-tuning or recalibration. Improvements also hold across three VLA backbones and separately trained real-world policies, where mean success increases from 71.0% to 83.0%. Component ablations and timing interventions support the value of explicit grasp-intent conditioning and expressive hand generation. These results identify arm-hand coordination as an actionable bottleneck in adapting parallel-gripper priors to dexterous manipulation.

cs.RO↗

A semi-implicit unified gas-kinetic wave-particle method with energy-conserving electromagnetic coupling for multiscale plasmas

In multiscale plasmas, kinetic regions coexist with collision-dominated, strongly magnetized and quasi-neutral ones. In these regions, the collision time, the cyclotron period and the plasma period, respectively, are far shorter than the flow time scales. Schemes that treat the collisions and the field coupling explicitly must take time steps below these times and below the light-crossing time of a cell. This paper presents a semi-implicit unified gas-kinetic wave-particle (UGKWP) method for the multispecies Vlasov-Maxwell system with Bhatnagar-Gross-Krook collisions. The electric field follows from a reformulated Maxwell equation, a single linear system with a unique solution for any time step. Its current response is built from the implicit-midpoint velocity map, which then pushes the particles and the hydrodynamic wave alike. The field step conserves the discrete total energy exactly and remains stable without resolving the cyclotron period, the plasma period or the light-crossing time, so that the time step follows from the flow and sound speeds. A Gauss's law correction adjusts the electric field and the joint charge of particles and wave while conserving the mass, momentum and energy of every charged species. The method reproduces whistler waves and electron plasma waves from Landau damping to collision-dominated oscillations. It agrees with particle-in-cell, two-fluid and magnetohydrodynamic references in a Brio-Wu problem from weak to strong collisionality and magnetization, the Orszag-Tang vortex and collisionless magnetic reconnection. It needs far less CPU time than either the UGKWP method with an explicit solver for the perfectly hyperbolic Maxwell equations or a deterministic multiscale kinetic scheme. In collisional plasmas, it also needs less CPU time than an implicit particle-in-cell method with the same collision operator, increasingly so as collisions strengthen.

physics.plasm-ph↗

Learning from Next-Frame Prediction: Towards Unified and Scalable Visual World Modeling

Video provides a natural source of supervision: what happens next serves as a learning signal for visual world modeling. Yet masked video pretraining focuses on recovering missing content from surrounding context rather than predicting the future from the past. Autoregressive methods instead predict forward one token at a time, but have yet to match joint-embedding methods in representation quality. This gap motivates us to revisit the granularity and space of future prediction. In this work, we introduce VANE, a self-supervised visual pretraining framework built on next-frame prediction in embedding space. A causal encoder summarizes past observations for a state predictor that infers the representation of the entire next frame, making temporal evolution the primary learning signal. This prediction is carried out in a fixed embedding space provided by a frozen semantic tokenizer, emphasizing scene content and structure. To retain fine-grained spatial detail alongside semantic content, a lightweight auxiliary generative decoder reconstructs future-frame patches conditioned on the predicted representations. To unify visual modeling across images and videos, we construct a local-to-global crop sequence for still images, allowing the same encoder to predict newly revealed regions from earlier crops under the same objective. Experiments show that VANE improves with model and data scaling, achieving SOTA on most anticipation, recognition, and dense perception benchmarks among unified encoders. Especially on EPIC-KITCHENS-100 action anticipation, it surpasses V-JEPA 2.1 by 7.1 points in mean-class recall@5. These results demonstrate that next-frame prediction in embedding space is effective and scalable for visual world modeling.

cs.CV↗

Optimal galaxy magnification: generalized fundamental-plane residuals and number counts are all you need

Weak-lensing magnification complements cosmic shear through its effects on galaxy fluxes, sizes, and number counts. We derive the optimal magnification statistics from galaxy properties and show that the optimal measurement can be decomposed into an optimal mean statistic and total number counts. The optimal mean statistic can be constructed from the residual distribution of a generalized fundamental plane (GFP), $F(i,\boldsymbol{X})$, which predicts galaxy size or, equivalently, flux from surface brightness $i$ and additional lensing-invariant properties $\boldsymbol{X}$. The function $F$ can take any form that reduces the intrinsic scatter and thereby improves magnification precision. For a Gaussian residual distribution, the optimal mean statistic reduces to the inverse-variance-weighted residual, or simply the residual for constant scatter. Together with total number counts, this statistic captures all the magnification Fisher information under Poisson statistics, achieving the same precision as a full-likelihood analysis. This provides a simple and physically intuitive optimal compression of the full magnification likelihood, avoiding the need to model the potentially high-dimensional galaxy-property distribution, particularly when additional observables are incorporated to reduce the GFP scatter. Our results demonstrate that GFP residuals and number counts provide an optimal combination for weak-lensing magnification measurements in galaxy surveys.

astro-ph.CO↗

A nonequilibrium-capturing unified gas-kinetic scheme I: discrete orthogonal basis construction and expansion on velocity space

Velocity space discretization is a primary source of computational cost in deterministic methods for multiscale gas flows. A fixed velocity grid often retains redundant kinetic degrees of freedom (DOFs), particularly under near equilibrium conditions. To address this issue, this work develops a nonequilibrium-capturing formulation based on the unified gas kinetic scheme (UGKS), in which the number of kinetic DOFs adapts locally to the degree of nonequilibrium. A discrete orthogonal basis is constructed for numerical implementation, and the velocity distribution function is decomposed into an analytical local equilibrium and a basis expanded nonequilibrium component. The retained, locally adapted DOFs are evolved directly, substantially lowering computational overhead while preserving the discrete conservation of mass, momentum, and energy, as well as accurately capturing nonequilibrium states. The formulation retains the core multiscale coupling of molecular transport and collision inherent to UGKS, providing explicit and implicit schemes for unsteady and steady flows, respectively. Numerical tests across a wide range of flow speeds and rarefaction regimes (from continuum to highly rarefied) demonstrate close agreement with benchmark solutions. For unsteady flows, the explicit scheme achieves speedups of 12.64 to 58.50 over the standard UGKS. In steady state tests, only about $2\%$ of the original kinetic DOFs are required in continuum regions, accelerating each implicit iteration by a factor of 33.87. Even in the most rarefied cases, the scheme captures nonequilibrium behavior using only about $27\%$ of the original DOFs.

physics.comp-ph↗

An implicit multiscale solver for steady kinetic equations based on wave-particle decomposition

Steady kinetic computations require accurate transport on meshes that do not resolve the mean free path and efficient removal of slowly decaying macroscopic errors. We develop a closed wave-particle decomposition (WPD) and a residual-based implicit solver for this problem. The wave accumulates only the Maxwellian along a local characteristic horizon and is determined by the conservative variables; the particle carries the kinetic complement and collision remainder. Under moment compatibility, the closed system is equivalent to the kinetic equation. For the Bhatnagar-Gross-Krook and Shakhov models, a hydrodynamic expansion gives the wave and particle contributions to the Euler and Navier-Stokes fluxes. The implicit iteration couples macroscopic prediction to microscopic correction through a discrete endpoint term and synchronizes the conservative moments after each accepted correction. Its stationary states satisfy the target discrete kinetic equation independently of the iteration controls. Formal analysis gives the smooth fixed-mesh fluid limit. Linear model analysis relates convergence to the horizon-to-relaxation ratio, incomplete macroscopic solves and explicit collision remainders. One- and two-dimensional computations of a normal shock, internal flows and hypersonic external flows test these properties across rarefied and continuum regimes. The results show kinetic fidelity and substantial reductions in iteration counts and wall time relative to the conventional iterative scheme near the continuum limit. Collision-model comparisons and parameter studies support the predicted convergence mechanisms.

physics.comp-ph↗

Harmonic Balance Unified Gas-Kinetic Scheme for Multiscale Periodic Flows

A time-domain harmonic balance unified gas-kinetic scheme (HB-UGKS) is developed for simulating periodic non-equilibrium flows across all Knudsen regimes. By applying a time-spectral operator, the unsteady periodic problem is reformulated into a block-coupled, quasi-steady system. To preserve the intrinsic multiscale transportcollision flux coupling of the UGKS, a source-separated formulation incorporates the harmonic balance coupling strictly at the cell-residual level. This allows the periodic limit cycle to be resolved directly via pseudo-time marching with local time-stepping, advancing all temporal collocation points simultaneously and completely bypassing physical startup transients. The framework is validated against three complementary benchmarks: two small-amplitude sheardriven oscillatory flows and a thermally driven cavity under finite-amplitude excitation. The scheme accurately captures intricate non-equilibrium kinetic phenomena, including aspectratio anti-resonance scaling, dynamic shear traction, nonlinear waveform distortions, and acoustic streaming. Across all cases, the HB-UGKS preserves the fidelity of standard timeaccurate simulations, while achieving order-of-magnitude speedups in high-frequency regimes where long transient timescales dominate.

physics.flu-dyn↗

MORPH: Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization

Embedding-based retrieval typically returns highest-scoring items, but many production scenarios require items that satisfy a target attribute while preserving a fine-grained pattern expressed by seed examples. We formalize this as pattern-preserving attribute retrieval. Standard approaches fail: averaging seeds preserves the pattern but misses the attribute; global attribute retrieval drifts to unrelated patterns. We approach the task with continuous generative retrieval, where a model reads item-embedding sequences and generates query embeddings for nearest-neighbor search. We propose MORPH: Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization, a staged framework with large-scale raw-sequence pretraining, Metric-Ordered Sequence (MOS) training, and final HPPO alignment. MOS construction turns sparse online metric labels into in-pattern trajectories; MOS CPT/SFT then trains the generator through shared multi-domain continuation pretraining and domain-specific tail-centroid supervised fine-tuning. HPPO uses a hybrid pool of static and policy-generated candidate embeddings, labels them with true online intersection metrics, applies iterated preference optimization, and employs a Pareto pair filter to exclude winners that lower pattern purity. Across four large-scale attribute domains under strict item- and pattern-holdout protocols, MOS training improves the primary intersection metric over a strong pretrained generative retriever in every domain-split cell, and the complete Pareto-filtered HPPO procedure improves it further, with paired-bootstrap-significant gains on seven of the eight cells - the exception being the D4 pattern-holdout split. Ablations confirm that the Pareto pair filter improves the attribute-pattern tradeoff on D1-D3, and that hybrid static/policy candidates are complementary.

cs.AI↗

Loop Dropout: Regularizing Shared Updates in Looped Language Models

Looped language models separate computational depth from parameter count by repeatedly applying the same transformer block. Adapting these models requires a shared update that remains effective as hidden states evolve throughout the recurrent computation. Our empirical analysis reveals a pronounced late-loop bias in standard low-rank adaptation (LoRA): the shared update provides limited adaptation at early loop positions. This imbalance motivates training shared updates under varying combinations of their applications. Randomly omitting adapter applications alone, however, does not improve task performance; it reduces expected update strength during training while leaving inference unchanged. We introduce Loop Dropout, which couples stochastic masking of adapter applications with inverse-survival rescaling to preserve expected update strength and promote effective adaptation across loops. Extensive experiments demonstrate improved mathematical reasoning across model sizes, adapter ranks and training recipes, with benefits extending to general instruction tuning and code generation. Loop Dropout outperforms existing LoRA variants and adapter regularizers, while further analysis shows stronger early-loop adaptation. Every backbone loop remains active, and inference applies the adapter at all loops using standard LoRA without additional trainable parameters or inference computation. Code is available at https://github.com/NUS-HPC-AI-Lab/loop-dropout .

cs.LG↗

A wave-particle decomposition framework for multiscale kinetic transport: continuous-spectrum equations and coupled wave-particle iteration

The wave-particle decomposition is presented as a framework for multiscale kinetic transport, comprising a theory and an algorithm. Along the characteristics of the kinetic equation, the equilibrium integral of the solution over a local kinetic horizon defines a wave, which depends only on the conservative variables, and its exact complement defines a particle. Both obey kinetic equations whose moments are extended Navier-Stokes equations, and the conservation laws together with the particle equation form a closed system that is equivalent to the kinetic equation for every horizon. Parametrized by the ratio of the horizon to the relaxation time, these systems form a continuous spectrum from kinetic theory to Navier-Stokes hydrodynamics that adapts to the flow from cell to cell. The algorithm is a coupled wave-particle iteration for steady flows. It alternates a macroscopic iteration for the conservative variables, with the particle frozen, and a kinetic iteration for the particle, with the wave fixed. Its fixed point is the discrete kinetic solution, it is asymptotic preserving, and its near-continuum convergence factor is bounded by the share of the transport carried by the particle, a share that decays exponentially with the horizon-to-relaxation ratio. The collision model enters only through a conservative remainder, and the equations and the iteration are written out for the Bhatnagar-Gross-Krook, Shakhov and ellipsoidal statistical models, the Boltzmann and Landau equations, neutron transport and radiative transfer. The decomposition is related to the unified gas-kinetic wave-particle method, the micro-macro decomposition, penalization and synthetic acceleration. Computations of hypersonic flows past a cylinder and around the three-dimensional X38 vehicle illustrate how the division of the distribution adapts to the flow and how strongly the iteration is accelerated.

physics.flu-dyn↗

Wave-Particle Decomposition for Kinetic Equations II: Full Boltzmann equation

The wave-particle decomposition of kinetic relaxation equations is extended to the full Boltzmann equation. An identity transformation writes the collision term as a relaxation toward the local Maxwellian plus a conservative remainder. Along characteristics, the equilibrium integral over a local kinetic horizon defines the wave, and its exact complement, the particle, carries the remainder. The moments of the wave and particle equations are extended Navier-Stokes equations, and the total conservation law and the particle equation form the wave-particle multiscale equations, a closed system, exact for every horizon, that spans a continuous spectrum from the Boltzmann equation to Navier-Stokes hydrodynamics, parametrized by the horizon-to-relaxation ratio. A coupled implicit wave-particle iteration solves this system, updating the conservative variables with the particle fixed and the particle with the predicted wave and an endpoint traction from the macroscopic prediction. Under discrete compatibility assumptions the iteration preserves the discrete Boltzmann solution as its fixed point, and its near-continuum convergence factor is bounded by a particle transport share decaying exponentially with this ratio. Tests on a normal shock, Couette and cavity flows, and hypersonic cylinder and Apollo flows agree with the conventional iterative scheme in rarefied and transition regimes and with Navier-Stokes solutions near the continuum limit. There the wave-particle iteration needs about three to more than four orders of magnitude fewer iterations than this scheme or, where it did not converge, its Shakhov version, and the Boltzmann solution costs one to four times the wall time of a Navier-Stokes solver on the same mesh.

physics.flu-dyn↗

Learning from Next-Frame Prediction: Autoregressive Video Modeling Encodes Effective Representations

Recent advances in pretraining general foundation models have significantly improved performance across diverse downstream tasks. While autoregressive (AR) generative models like GPT have revolutionized NLP, most visual generative pretraining methods still rely on BERT-style masked modeling, which often disregards the temporal information essential for video analysis. The few existing autoregressive visual pretraining methods suffer from issues such as inaccurate semantic localization and poor generation quality, leading to poor semantics. In this work, we propose NExT-Vid, a novel autoregressive visual generative pretraining framework that utilizes masked next-frame prediction to jointly model images and videos. NExT-Vid introduces a context-isolated autoregressive predictor to decouple semantic representation from target decoding, and a conditioned flow-matching decoder to enhance generation quality and diversity. Through context-isolated flow-matching pretraining, our approach achieves strong representations. Extensive experiments on large-scale pretrained models demonstrate that our proposed method consistently outperforms previous generative pretraining methods for visual representation learning via attentive probing in downstream classification.

cs.CV↗

Multiscale Kinetic Methods for Nonequilibrium Flow and Transport

Nonequilibrium flow and transport problems are inherently multiscale. Kinetic theory provides a fundamental physical basis for describing such phenomena, since it connects microscopic transport and interaction processes with emergent macroscopic behavior across regimes. In many situations, however, continuum descriptions lose validity in parts of the domain, whereas fully resolved kinetic descriptions become prohibitively expensive when numerical resolution remains tied to the smallest collision scales. Over the past two decades, a broad class of multiscale kinetic methods has therefore been developed to bridge rarefied, transitional, and continuum regimes in gas dynamics and other carrier-based transport systems. Existing reviews have clarified important parts of this field, including general numerical methods for kinetic equations, asymptotic-preserving methodology, and specific method families. This review adopts a different perspective by examining the subject through four interacting layers: numerical methods, computational strategies, asymptotic properties, and framework-level formulations. It surveys the main developments along these lines and emphasizes the common principles that connect them, including transport-interaction coupling, asymptotic consistency, scale-adaptive representation, and scale-dependent physical description. From this perspective, multiscale kinetic computation has evolved into a broader transport methodology for nonequilibrium systems across scales.

physics.flu-dyn↗

A polar-harmonic unified gas-kinetic scheme for magnetized ion dynamics from cyclotron kinetics to the Hall-Pedersen constitutive limit

Magnetized ion transport in weakly ionized plasmas ranges from gyroangle-dependent kinetics to Hall-Pedersen drift-diffusion as collisionality and magnetization vary. We develop a polar-harmonic unified gas-kinetic scheme (PH-UGKS) for the ion Vlasov-BGK equation in a uniform magnetic field. The scheme evolves the full ion distribution by coupling a conservative density update to exponential evolution of its nonequilibrium component. Exact collision-rotation integration in gyroangle Fourier space is combined with a time-averaged kinetic flux that incorporates spatial transport and electric acceleration, together with a compact Hall-Pedersen correction to the density flux. The scheme conserves ion number, and analysis establishes second-order temporal consistency and asymptotic preservation of the Hall-Pedersen density limit at fixed magnetization. Numerical tests reproduce ion Bernstein dispersion and Dory-Guest-Harris growth rates and resolve changes in the gyroharmonic spectrum as the collision-to-gyrofrequency ratio varies. The driven ion-flux response agrees with an independent characteristic-Volterra reference, including finite-frequency departures from the instantaneous Hall-Pedersen relation. In collisional tests, accurate responses are obtained with time steps far larger than both the collision time and the gyroperiod. Fixed-resolution density tests confirm convergence to the corresponding Hall-Pedersen discretization. The same kinetic formulation thus connects kinetic response and macroscopic transport without switching to a fluid solver or subcycling microscopic time scales.

physics.plasm-ph↗

Sizing the Universe with DESI Galaxy Sizes: Plain Fundamentals of Fundamental-Plane Lensing

Weak gravitational lensing provides a powerful way to map cosmic structure, but most current measurements rely on galaxy shape distortions from deep imaging surveys and are affected by systematics such as intrinsic alignments, photometric-redshift uncertainties and shape-measurement biases. Here we present a spectroscopic galaxy-galaxy lensing magnification measurement using Fundamental-Plane (FP) size residuals, $δ_r\equivΔ\log_{10}R_\mathrm{e}$, from 3.26 million DESI luminous red galaxy (LRG) sources behind DESI Bright Galaxy Survey lenses. The FP-like relation predicts the intrinsic sizes of LRGs from lensing-invariant quantities, including velocity dispersion $σ_0$ and surface brightness $I_\mathrm{e}$, with a scatter of about 0.06-0.07 dex. Lensing magnifies LRG sizes, giving the direct convergence response $δ_r(κ)=κ/\ln 10$, but we show that the full lensing response is modified by magnification bias, because fitting $R_\mathrm{e}$ with $I_\mathrm{e}$ inevitably induces a magnitude dependence in $\barδ_r(m)$. After calibrating this response, we recover surface-density profiles with uncertainties comparable to those from individual Stage-III shear surveys using 5-20 million higher-redshift sources. The corresponding excess surface-density profiles agree with shear-based measurements. We further show that the estimator is robust to size-measurement uncertainties, with a convergence multiplicative bias only $\simeq -0.2$ times the size bias. FP lensing therefore provides a clean, spectroscopic and complementary probe of cosmic structure.

astro-ph.CO↗

Concept-Level Risk and Calibration for Governance in Diffusion Foundation Models

Diffusion models have become a core paradigm for multimedia generation, offering powerful concept-driven controllability for personalization, semantic editing, and selective unlearning. However, as semantic control extends beyond natural-language prompts to learned embeddings and intervention pipelines, the safety and governance of these systems become increasingly difficult to evaluate in a unified manner, especially for safety-sensitive, identity-linked, and other privacy-relevant concepts. Existing studies mainly rely on heuristic audits, adversarial probing, or task-specific erasure benchmarks, and therefore provide limited support for systematic comparison across models, conditioning channels, and deployment conditions. We present a concept-level probabilistic audit and reporting framework for diffusion models. We formalize governance-relevant concept behaviors as Bernoulli semantic events induced by stochastic generation, and define a Concept Risk Operator that maps model-channel configurations to structured risk profiles, enabling comparison across prompting interfaces, learned embedding channels, models, and recorded conditions. We apply sample-level post-hoc calibration and configuration-level risk aggregation, and show that probability error can change thresholded actions near policy boundaries. Experiments on SD1.5, SD2.1, and SDXL reveal consistent yet non-uniform operational risk patterns across concept families, channels, recorded conditions, and shifted protocols. In particular, embedding-based access and obfuscated prompts expose risks often understated by standard-prompt evaluation. A pooled multi-protocol calibrator improves held-out probability reliability, but we do not claim transfer from a standard-only calibrator. CLRC provides a common audit schema for probabilistic and decision-aware governance of multimedia generation systems.

cs.MM↗