Searcharxiv⌕ Search

arXiv subjects

Yu Liang

Publications and source records attributed to Yu Liang.

At least 19 recordsLinked to original sources
arXiv2026

Coherent multipath wave response on Reissner-Nordstr\"{o}m analogue surface

To uncover how the intrinsic metric of a relativistic compact object governs macroscopic wave phenomena, we establish a theoretical framework mapping the charge dependent spatial geometry of a Reissner-Nordstr\"{o}m (RN) black hole onto the coherent response of an analogue curved surface. By solving an exact spatial geodesic boundary value problem on an isometrically embedded equatorial slice, we extract the discrete multi-loop path-length spectrum and convert this geometric backbone into a physical wave field via a finite-path surface Huygens-Fresnel construction. We analytically compute the arbitrary order winding trajectories alongside their high winding accumulation limits, demonstrating that the analogue charge acts as a precise physical dial that reconfigures the event horizon throat and fundamentally reorganizes the discrete path sequence. Furthermore, we find that this underlying geometric deformation uniquely dictates the macroscopic interference, revealing that steady state spatial fringes, spectral resonance combs, and transient temporal echo ladders are intrinsically unified physical projections of a single charge controlled path spectrum. This systematic parameter to response methodology establishes a rigorous theoretical bridge between strong field gravitational lensing and tabletop transformation optics, providing a highly tunable blueprint for future multi domain analogue gravity experiments.

gr-qc↗
arXiv2026

Stationary Dirac condensates around Kerr black holes

Ultralight bosonic fields can form macroscopic clouds around rotating black holes, whereas the existence of analogous stationary fermionic condensates is strictly constrained by their intrinsic spin. Here we establish a complete geometric and kinematic framework to resolve the stationary bound states of massive Dirac fields on Kerr and Kerr-Newman backgrounds. By mapping the Kerr-Dirac system to a globally integrated sourced radial problem, we strictly isolate the boundary constraints dictated by horizon causality. The angular sector reveals a fundamental topological distinction: because the azimuthal quantum number is strictly half-integer, the regular boundary branches prevent the local field density from vanishing on the rotation axis. Consequently, rotating fermionic clouds inherently form globally filled, oblate geometries, in stark contrast to the hollow toroidal structures characteristic of scalar condensates. Crucially, our radial indicial analysis unveils the exact mathematical origin of the absence of synchronized Dirac hair. Precisely at the kinematic synchronization locus, the Frobenius matrix of the Dirac operator is non-defective and entirely devoid of logarithmic divergences. Without these singular branches to be selectively excised by boundary regularity, the physical burden of existence falls entirely onto the causal flux barrier, which strictly trivializes the zero-source amplitude. This synchronization veto demonstrates that a black hole's capacity to support macroscopic stationary fields is governed not merely by superradiant kinematics, but by the profound interplay between local horizon causality and quantum spin statistics.

gr-qc↗
arXiv2026

PathView-Bench: Can Multimodal Large Language Models Achieve Fine-grained Multiscale Understanding of Pathology Images?

Multimodal large language models (MLLMs) are increasingly used to analyze pathology images. However, dominant multimodal benchmarks in pathology mainly score final diagnostic answers, captions, or reports. These evaluations provide limited insight into whether a model understands the multiscale visual content needed for pathology reasoning and decision-making. We introduce PathVU, a vision-anchored benchmark for fine-grained and multiscale visual understanding in computational pathology. Built from 23 public pathology imaging datasets with human-supervised labels and spatial annotations, PathVU evaluates MLLM understanding in two fields of view: Region FOV for high-resolution local regions and Slide FOV for macro whole-slide views. By converting raw annotations into deterministic task targets, PathVU enables programmatic scoring of region localization, visual recognition, quantity estimation, spatial reasoning, and insufficient-context judgment. The benchmark contains 14 VQA-style tasks, 61,673 images, and 308,070 samples across 28 organs and 7,253,526 annotations. Evaluating 18 representative general-purpose, medical-domain, and pathology-oriented MLLMs, we observe substantial limitations even in advanced models on fine-grained visual tasks across multiscale pathology images. PathVU provides a reproducible basis for developing and evaluating pathology MLLMs with explicit multiscale visual understanding.

cs.AI↗
arXiv2026

SuperPass: Fast-Tracking Blocking Threads to Mitigate Priority Inversion on Mobile Devices

Priority inversion occurs when a high-priority thread is delayed by a lower-priority one. Although well studied in real-time systems, its impact in general-purpose OSes (e.g., Android) remains underexplored. On Android, we find that priority inversions happen frequently and can delay latency-critical threads, degrading user experience. For example, the foreground app's UI thread is frequently blocked by low-priority threads, with blocking durations of up to 210 ms, enough to cause dropped frames. Existing solutions designed for real-time systems fail to eliminate long priority-inversion blockings on latency-critical threads and may introduce high overhead on Android. To solve this problem, we uncover two insights on Android: 1) long blockings are mainly due to the accumulated CPU waiting time of low-priority blocking threads rather than their critical-section latency; and 2) although latency-critical threads can be blocked by many concurrent readers, tracking a limited number of them is sufficient to achieve good responsiveness with low overhead in most cases. Guided by these insights, we propose SuperPass, a lightweight kernel mechanism that mitigates priority inversion by fast-track scheduling of low-priority threads blocking latency-critical threads. It introduces a scheduler fast track that grants immediate CPU access to threads blocking latency-critical threads, and employs a lock-level detector that effectively identifies most such blocking threads. We evaluate SuperPass on a Google Pixel 8 smartphone. Taking UI thread as a case study, SuperPass decreases the 99.9th-percentile blocking duration by 72.0% and blocking count by 47.7% on average compared to the default scheduler, and reduces janky frames by 29.2% with a system-wide CPU overhead of only 0.74%. SuperPass also outperforms existing approaches including priority inheritance, real-time UI promotion, and Proxy Execution.

cs.OS↗
arXiv2026

Optical Appearances of Accreting Ellis-Bronnikov Wormholes Observed from Both Sides of Throats

This study investigates the optical appearance of the Ellis-Bronnikov wormhole as viewed from both sides of its throat, under conditions of optically thick and thin accretion. By solving the geodesic equation, we derive the relationship between the impact parameter and the aiming distance of photons, and found that if the observer and the accretion disk are located on both sides of the throat, these two quantities are not equal. The optical image of the wormhole observed from the other side of the throat is obtained through the ray-tracing method. For optically thick accretion, increases in the parameter $n$ lead to an increase in the apparent size of the wormhole but a decrease in its brightness. For optically thin accretion, the image is similar to the internal and external inversion of the image observed from the other side. Furthermore, for optically thin accretion flows, the direct image does not block the emission from higher-order images, allowing radiation emitted from regions much closer to the event horizon to reach the observer. Our simulation results show that when the observer is on the $\mathcal{R}^+$ side, EB wormholes with small $n$ can mimic the images taken by the EHT to some extent, while wormholes with large $n$ or with the observer on the $\mathcal{R}^-$ side can be ruled out.

gr-qc↗
arXiv2026

Gravitational waveforms from periodic orbits around Gauss-Bonnet black holes

Extreme mass-ratio inspirals (EMRIs) constitute one of the most promising probes of strong field gravity for future space borne gravitational-wave observatories. As a representative higher-curvature extension of General Relativity (GR), four-dimensional Einstein-Gauss-Bonnet (4D EGB) gravity is distinguished by its strictly linear geometric coupling. By this mathematical property, the pathological Fisher-matrix singularities that typically plague conventional modified black hole models are effectively evaded, thereby providing an ideal framework to test topological deviations from classical spacetimes. Through the classification of equatorial periodic orbits via an integer taxonomy $(z,w,v)$, it is demonstrated that even modest Gauss-Bonnet couplings ($\alpha \sim 0.1M^2$) imprint measurable geometric signatures onto the zoom-whirl architecture. Although the global conservative energy budget is shifted by a mere $\sim 0.2\%$, the short-range repulsive EGB core severely alters the strong field whirl dynamics, whereby a resolvable macroscopic dephasing of several radians per orbit is accumulated. Through semi-relativistic waveform modeling, it is revealed that this temporal compression manifests as a rigid, high-frequency stretching of the gravitational-wave harmonic comb -- a clean, amplitude-independent spectral signature ideally suited for detection by LISA, Taiji, and TianQin. A rigorous Fisher information analysis confirms that for a typical four-year observation at a signal-to-noise ratio of $\rho=20$, the marginalized error on the EGB coupling can be tightly bounded to $\sigma_\alpha \sim \mathcal{O}(10^{-6}) M^2$, with virtually negligible parameter degeneracy with the orbital eccentricity.

gr-qc↗
arXiv2026

ScaleDisturb: Exploiting Temporal Asymmetry to Amplify Read Disturbance in Modern DRAM Chips

DRAM suffers from read disturbance phenomena (e.g., RowHammer and RowPress), where repeatedly accessing or continuously keeping open a DRAM row (aggressor row) induces bitflips in other physically nearby unaccessed rows (victim rows). The disturbance mechanism is practically exploitable from the software stack and worsens across generations with continued density scaling. DRAM read disturbance is highly sensitive to memory access patterns, yet prior work explores read disturbance under only a limited set of access patterns. We present ScaleDisturb, a new DRAM access pattern that can amplify DRAM read disturbance by asymmetrically extending the open time of two aggressor rows. Our rigorous experimental characterization of 196 DDR4 and 3 HBM2 DRAM chips shows that ScaleDisturb (1) leads to bitflips at significantly fewer row activations, compared to state-of-the-art memory access patterns, (2) makes read disturbance attacks easier across all tested DRAM chips, (3) increases DRAM vulnerability to read disturbance as DRAM manufacturing technology scales down to smaller node sizes. We showcase a proof-of-concept attack on a real system where a user-level program leveraging ScaleDisturb induces more bitflips than state-of-the-art RowHammer and RowPress memory access patterns. We describe and evaluate four solutions for mitigating read disturbance bitflips in the presence of ScaleDisturb and call for more research on the topic.

cs.CR↗
arXiv2026

RAISE: RAG Design as an Architecture Search Problem

Retrieval-augmented generation (RAG) systems expose numerous design choices spanning query rewriting, chunking, retrieval depth, reranking, and context compression. In practice, these choices are often configured through heuristics, hindering systematic evaluation and reproducibility across settings. We argue that this challenge is best formulated as RAG architecture search. To support controlled and reproducible study of this problem, we introduce the RAG Intelligence Search Engine (RAISE), a comprehensive framework and benchmark for RAG hyperparameter optimization, which evaluates optimization methods for RAG pipelines under standardized search spaces and budgets. RAISE implements 13 search algorithms and evaluates them across seven public text and multimodal datasets using three random seeds. Our experiments show that optimization performance is highly task-dependent: methods that perform strongly on one dataset may not generalize consistently across others, cautioning against interpreting aggregate rankings as evidence of universally superior strategies. RAISE provides a common experimental substrate for fair, reproducible, and systematic research on RAG hyperparameter optimization.

cs.AI↗
arXiv2026

Thinking as Compression: Your Reasoning Model is Secretly a Context Compressor

Context compression aims to shorten long context inputs with minimal information loss for LLM inference acceleration. While existing methods have shown promise, they typically rely on complex compression modules or compression-specific training, leaving the intrinsic capabilities of LLMs underexplored. In contrast, this work reveals that a thinking model itself can naturally compress long contexts by organizing task-relevant information. We thus derive Thinking as Compression (TaC), a new compression paradigm that treats thinking itself as compressed context. Without relying on specific dedicated compressor, TaC directly prompts the thinking model to generate thinking traces as the shortened context, already outperforming most representative compression methods. Further, given that raw thinking output may struggle with budget control and shortcut behaviors, we introduce Thinking as Compression Constrained (TaC-C), leveraging a simple reward-driven optimization framework to elicit intrinsic thinking as compact and controllable compressed context. Experiments across four long-context QA benchmarks demonstrate that TaC-C consistently outperforms existing baselines. At 4x and 8x compression ratios, it surpasses the strongest competitor by 17.4% and 23.4% in average F1, and by 15.7% and 21.7% in average Exact Match Score (EM), respectively.

cs.AI↗
arXiv2026

Development of a xenon triple point apparatus suitable for calibrating long-stem SPRTs and preliminary measurements of the temperature

Xenon is of high chemical-physical stability and health compatibility. The xenon triple point (Xe TP) is accounted for a promising candidate replacing the mercury triple point (Hg TP) from the set of the defining fixed points of the international temperature scale ITS-90. The success of the alternative highly depends on the level of the realization of the Xe TP using long-stem standard platinum resistance thermometers (LSPRTs). We report in this article our study on the development of an immersion-type Xe TP apparatus, which is suitable for calibration of both LSPRTs and capsule standard platinum resistance thermometers (CSPRT). We realize the melting plateaus of the Xe TP using the continuous heating method on the apparatus. The effective melting plateaus extend for 8-12 hours long with temperature flatness range of 0.37 mK-1.0 mK over the melted fractions from 0.2 to 0.75. We find the axial heat leak contributing a principal effect influencing measurements of the Xe TP. We investigate the effect by varying the offset temperatures on the outer wall of the Xe TP cell. We measure the Xe TP using two LSPRTs upon correction of the axial heat leak. The new measurement, giving the Xe TP of 161.405 71 (55) K (k=1) at the melted fraction F=1.0, agrees well with those previously obtained by the adiabatic apparatuses. Their differences fall within 0.11 mK to 0.42 mK. by. Those differences are well covered by the estimated measurement uncertainty.

physics.ins-det↗
arXiv2026

COSMIC 1001: Engaging Future Speculation on Space Exploration with Generative AI

Cosmic 1001 is an interactive installation that transforms space exploration history into a speculative news experience. Participants first browse a news-based archive of major space events, then pose future-oriented questions or specify conditions such as year, celestial body, or mission name. In response, AI generates a future news item including a headline, article, narration, and visual media. These outputs are accumulated in the Future Tunnel, a shared visualization where individual stories form a collective landscape of possible futures. By combining historical space events with science fiction references, the installation explores a space between documentation and imagination, treating the future not as a fixed prediction but as a visible and discussable speculation.

cs.HC↗
arXiv2026

ConsistRM: Improving Generative Reward Models via Consistency-Aware Self-Training

Generative reward models (GRMs) have emerged as a promising approach for aligning Large Language Models (LLMs) with human preferences by offering greater representational capacity and flexibility than traditional scalar reward models. However, GRMs face two major challenges: reliance on costly human-annotated data restricts scalability, and self-training approaches often suffer from instability and vulnerability to reward hacking. To address these issues, we propose ConsistRM, a self-training framework that enables effective and stable GRM training without human annotations. ConsistRM incorporates the Consistency-Aware Answer Reward, which produces reliable pseudo-labels with temporal consistency, thereby providing more stable model optimization. Moreover, the Consistency-Aware Critique Reward is introduced to assess semantic consistency across multiple critiques and allocates fine-grained and differentiated rewards. Experiments on five benchmark datasets across four base models demonstrate that ConsistRM outperforms vanilla Reinforcement Fine-Tuning (RFT) by an average of 1.5%. Further analysis shows that ConsistRM enhances output consistency and mitigates position bias caused by input order, highlighting the effectiveness of consistency-aware rewards in improving GRMs. Our implementation is available at https://github.com/yuliangCarmelo/ConsistRM.

cs.AI↗
arXiv2026

ReflectRM: Boosting Generative Reward Models via Self-Reflection within a Unified Judgment Framework

Reward Models (RMs) are critical components in the Reinforcement Learning from Human Feedback (RLHF) pipeline, directly determining the alignment quality of Large Language Models (LLMs). Recently, Generative Reward Models (GRMs) have emerged as a superior paradigm, offering higher interpretability and stronger generalization than traditional scalar RMs. However, existing methods for GRMs focus primarily on outcome-level supervision, neglecting analytical process quality, which constrains their potential. To address this, we propose ReflectRM, a novel GRM that leverages self-reflection to assess analytical quality and enhance preference modeling. ReflectRM is trained under a unified generative framework for joint modeling of response preference and analysis preference. During inference, we use its self-reflection capability to identify the most reliable analysis, from which the final preference prediction is derived. Experiments across four benchmarks show that ReflectRM consistently improves performance, achieving an average accuracy gain of +3.7 on Qwen3-4B. Further experiments confirm that response preference and analysis preference are mutually reinforcing. Notably, ReflectRM substantially mitigates positional bias, yielding +10.2 improvement compared with leading GRMs and establishing itself as a more stable evaluator. Our code is available at https://github.com/yuliangCarmelo/ReflectRM.

cs.AI↗
arXiv2026

Pedestrian Crossing Intent Prediction via Psychological Features and Transformer Fusion

Pedestrian intention prediction needs to be accurate for autonomous vehicles to navigate safely in urban environments. We present a lightweight, socially informed architecture for pedestrian intention prediction. It fuses four behavioral streams (attention, position, situation, and interaction) using highway encoders, a compact 4-token Transformer, and global self-attention pooling. To quantify uncertainty, we incorporate two complementary heads: a variational bottleneck whose KL divergence captures epistemic uncertainty, and a Mahalanobis distance detector that identifies distributional shift. Together, these components yield calibrated probabilities and actionable risk scores without compromising efficiency. On the PSI 1.0 benchmark, our model outperforms recent vision language models by achieving 0.9 F1, 0.94 AUC-ROC, and 0.78 MCC by using only structured, interpretable features. On the more diverse PSI 2.0 dataset, where, to the best of our knowledge, no prior results exist, we establish a strong initial baseline of 0.78 F1 and 0.79 AUC-ROC. Selective prediction based on Mahalanobis scores increases test accuracy by up to 0.4 percentage points at 80% coverage. Qualitative attention heatmaps further show how the model shifts its cross-stream focus under ambiguity. The proposed approach is modality-agnostic, easy to integrate with vision language pipelines, and suitable for risk-aware intent prediction on resource-constrained platforms.

cs.CV↗
arXiv2026

Robust Spiking Neural Networks Against Adversarial Attacks

Spiking Neural Networks (SNNs) represent a promising paradigm for energy-efficient neuromorphic computing due to their bio-plausible and spike-driven characteristics. However, the robustness of SNNs in complex adversarial environments remains significantly constrained. In this study, we theoretically demonstrate that those threshold-neighboring spiking neurons are the key factors limiting the robustness of directly trained SNNs. We find that these neurons set the upper limits for the maximum potential strength of adversarial attacks and are prone to state-flipping under minor disturbances. To address this challenge, we propose a Threshold Guarding Optimization (TGO) method, which comprises two key aspects. First, we incorporate additional constraints into the loss function to move neurons' membrane potentials away from their thresholds. It increases SNNs' gradient sparsity, thereby reducing the theoretical upper bound of adversarial attacks. Second, we introduce noisy spiking neurons to transition the neuronal firing mechanism from deterministic to probabilistic, decreasing their state-flipping probability due to minor disturbances. Extensive experiments conducted in standard adversarial scenarios prove that our method significantly enhances the robustness of directly trained SNNs. These findings pave the way for advancing more reliable and secure neuromorphic computing in real-world applications.

cs.CV↗
arXiv2026

What does a regular star look like?

Recently, astronomers discovered unusual Einstein cross images of the galaxy HerS-3, which feature a bright central spot. Motivated by studies of images produced by regular stars, it has been proposed that optical appearances caused by compact stars acting as gravitational lenses may account for this central bright spot. We further suggest that images produced by regular stars exhibit additional characteristics distinct from those of ordinary black holes, such as the possible partial or complete absence of secondary images. These phenomena may serve as favorable observational criteria for identifying regular stars in future searches.

gr-qc↗
arXiv2026

Advancing General-Purpose Reasoning Models with Modular Gradient Surgery

Reinforcement learning (RL) has played a central role in recent advances in large reasoning models (LRMs), yielding strong gains in verifiable and open-ended reasoning. However, training a single general-purpose LRM across diverse domains remains challenging due to pronounced domain heterogeneity. Through a systematic study of two widely used strategies, Sequential RL and Mixed RL, we find that both incur substantial cross-domain interference at the behavioral and gradient levels, resulting in limited overall gains. To address these challenges, we introduce **M**odular **G**radient **S**urgery (**MGS**), which resolves gradient conflicts at the module level within the transformer. When applied to Llama and Qwen models, MGS achieves average improvements of 4.3 (16.6\%) and 4.5 (11.1\%) points, respectively, over standard multi-task RL across three representative domains (math, general chat, and instruction following). Further analysis demonstrates that MGS remains effective under prolonged training. Overall, our study clarifies the sources of interference in multi-domain RL and presents an effective solution for training general-purpose LRMs.

cs.CL↗
arXiv2026

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Semantic ID (SID)-based recommendation is a promising paradigm for scaling sequential recommender systems, but existing methods largely follow a semantic-centric pipeline: item embeddings are learned from foundation models and discretized using generic quantization schemes. This design is misaligned with generative recommendation objectives: semantic embeddings are weakly coupled with collaborative prediction, and generic quantization is inefficient at reducing sequential uncertainty for autoregressive modeling. To address these, we propose ReSID, a recommendation-native, principled SID framework that rethinks representation learning and quantization from the perspective of information preservation and sequential predictability, without relying on LLMs. ReSID consists of two components: (i) Field-Aware Masked Auto-Encoding (FAMAE), which learns predictive-sufficient item representations from structured features, and (ii) Globally Aligned Orthogonal Quantization (GAOQ), which produces compact and predictable SID sequences by jointly reducing semantic ambiguity and prefix-conditional uncertainty. Theoretical analysis and extensive experiments across ten datasets show the effectiveness of ReSID. ReSID consistently outperforms strong sequential and SID-based generative baselines by an average of over 10%, while reducing tokenization cost by up to 122x. Code is available at https://github.com/FuCongResearchSquad/ReSID.

cs.IR↗