arXiv subjects

Peiyu Yang

Publications and source records attributed to Peiyu Yang.

At least 19 recordsLinked to original sources

arXiv2026

CliffRank: A Dual-Branch Framework for Activity-Cliff Ranking Prediction

Activity-cliff ranking remains difficult because local structural changes can cause large activity differences, while high-quality data that resolve the underlying mechanisms remain limited. To use available activity labels more effectively, we combine absolute-activity regression with ranking-consistency learning. CliffRank trains two parallel predictors with mean squared error, a thresholded listwise loss, and Pairwise Preference Consistency (PPC), which aligns relative ordering in the preference-probability space. On three antimicrobial peptide datasets, CliffRank with ESM2-t12 achieved the highest mean Spearman correlation of 0.5393 and mean Recall@50 of 21.4, although the leading method varied across individual datasets. On three small-molecule datasets, CliffRank with PNA, where PPC was activated after 120 epochs, achieved the highest mean Spearman correlation of 0.6890, while its mean Recall@50 of 30.4 matched that of ACANet-PNA. The PPC results also define its practical limits. Asymmetric initialization improved the MolCLR-GIN averages but did not improve every target. For PNA without pretrained weights, delayed PPC improved selected metrics, but no schedule was best for both mean Spearman correlation and mean Recall@50. Future work should evaluate more targets and antimicrobial peptide systems, develop adaptive PPC schedules, and incorporate protein or membrane context when available.

cs.LG↗

arXiv2026

Towards Efficient Large Language Model Serving: A Survey on System-Aware KV Cache Optimization

Despite the rapid advancements of large language models (LLMs), LLM serving systems remain memory-intensive and costly. The key-value (KV) cache, which stores KV tensors during autoregressive decoding, is crucial for enabling low-latency, high-throughput LLM inference serving. In this survey, we focus on system-aware KV infrastructure for serving LLMs (abbreviated as sKis). We revisit recent work from a system behavior perspective, organizing existing efforts into three dimensions: execution and scheduling (temporal), placement and migration (spatial), and representation and retention (structural). Furthermore, we analyze cross-behavior co-design affinity and behavior-objective links, highlighting future opportunities. Our work systematizes a rapidly evolving area, providing a foundation for understanding and innovating KV cache designs in modern LLM serving infrastructure.

cs.LG↗

arXiv2026

Semantic Robustness Certification for Vision-Language Models

Vision-language models (VLMs) are now widely used in downstream tasks. However, real-world applications often expose VLMs to distribution shifts induced by semantic variation (e.g., shape, size, and style). Robustness certification determines if a model's prediction changes when transformations are applied to its input. While most certification frameworks study geometric or pixel-level transformations over inputs, this work proposes a novel framework that enables certifying VLM robustness under semantic-level transformations. Leveraging the open-vocabulary capability of VLMs, we use text prompts as semantic proxies to construct transformations parameterized by an extent that controls the degree of semantic variation. By characterizing the VLM decision boundary in closed form, our framework quantitatively certifies extent intervals for which the predicted class remains unchanged under the semantic transformation. Our framework is the first to certify VLM robustness under semantic-level variations without requiring additional data for each variation, making it practical to apply. Experiments on both synthetic and real-world data show that our framework enables certifying robustness under diverse semantic variations across scenarios.

cs.LG↗

arXiv2026

MBD: A Model-Based Debiasing Framework Across User, Content, and Model Dimensions

Modern recommendation systems rank candidates by aggregating multiple behavioral signals through a value model. However, many commonly used signals are inherently affected by heterogeneous biases. For example, watch time naturally favors long-form content, loop rate favors short - form content, and comment probability favors videos over images. Such biases introduce two critical issues: (1) value model scores may be systematically misaligned with users' relative preferences - for instance, a seemingly low absolute like probability may represent exceptionally strong interest for a user who rarely engages; and (2) changes in value modeling rules can trigger abrupt and undesirable ecosystem shifts. In this work, we ask a fundamental question: can biased behavioral signals be systematically transformed into unbiased signals, under a user - defined notion of ``unbiasedness'', that are both personalized and adaptive? We propose a general, model-based debiasing (MBD) framework that addresses this challenge by augmenting it with distributional modeling. By conditioning on a flexible subset of features (partial feature set), we explicitly estimate the contextual mean and variance of the engagement distribution for arbitrary cohorts (e.g., specific video lengths or user regions) directly alongside the main prediction. This integration allows the framework to convert biased raw signals into unbiased representations, enabling the construction of higher-level, calibrated signals (such as percentiles or z - scores) suitable for the value model. Importantly, the definition of unbiasedness is flexible and controllable, allowing the system to adapt to different personalization objectives and modeling preferences. Crucially, this is implemented as a lightweight, built-in branch of the existing MTML ranking model, requiring no separate serving infrastructure.

cs.LG↗

arXiv2026

Towards Terrain-Aware Safe Locomotion for Quadrupedal Robots Using Proprioceptive Sensing

Achieving safe quadrupedal locomotion in real-world environments has attracted much attention in recent years. When walking over uneven terrain, achieving reliable estimation and realising safety-critical control based on the obtained information is still an open question. To address this challenge, especially for low-cost robots equipped solely with proprioceptive sensors (e.g., IMUs, joint encoders, and contact force sensors), this work first presents an estimation framework that generates a 2.5-D terrain map and extracts support plane parameters, which are then integrated into contact and state estimation. Then, we integrate this estimation framework into a safety-critical control pipeline by formulating control barrier functions that provide rigorous safety guarantees. Experiments demonstrate that the proposed terrain estimation method provides smooth terrain representations. Moreover, the coupled estimation framework of terrain, state, and contact reduces the mean absolute error of base position estimation by 64.8%, decreases the estimation variance by 47.2%, and improves the robustness of contact estimation compared to a decoupled framework. The terrain-informed CBFs integrate historical terrain information and current proprioceptive measurements to ensure global safety by keeping the robot out of hazardous areas and local safety by preventing body-terrain collision, relying solely on proprioceptive sensing.

cs.RO↗

arXiv2026

Attribution-Guided Model Rectification of Unreliable Neural Network Behaviors

The performance of neural network models deteriorates due to their unreliable behavior on non-robust features of corrupted samples. Owing to their opaque nature, rectifying models to address this problem often necessitates arduous data cleaning and model retraining, resulting in huge computational and manual overhead. In this work, we leverage rank-one model editing to establish an attribution-guided model rectification framework that effectively locates and corrects model unreliable behaviors. We first distinguish our rectification setting from existing model editing, yielding a formulation that corrects unreliable behavior while preserving model performance and reducing reliance on large budgets of cleansed samples. We further reveal a bottleneck of model rectifying arising from heterogeneous editability across layers. To target the primary source of misbehavior, we introduce an attribution-guided layer localization method that quantifies layer-wise editability and identifies the layer most responsible for unreliabilities. Extensive experiments demonstrate the effectiveness of our method in correcting unreliabilities observed for neural Trojans, spurious correlations and feature leakage. Our method shows remarkable performance by achieving its editing objective with as few as a single cleansed sample, which makes it appealing for practice.

cs.LG↗

arXiv2025

Quantum-elevated Chiral Discrimination for Bio-molecules

Chiral discrimination of enantiomeric biomolecules is vital in chemistry, biology, and medicine. Conventional methods, relying on circularly polarized light, face weak chiroptical signals and potential photodamage. Despite extensive efforts to improve sensitivity under low-photon exposure, classical chiral probes remain fundamentally bounded by the shot-noise limit due to quantum fluctuations. To beat these limitations, we demonstrate quantum-elevated chiral discrimination using continuous-variable polarization-entangled states as moderate-photon-flux, high-sensitivity, quantum-noise-squeezed chiral probes. We achieve a 5 dB improvement beyond the SNL in distinguishing L- and D-amino acids in liquid phase. This non-destructive, biocompatible protocol enables high-sensitivity chiral analysis, with broad implications for drug development, biochemical research, environmental monitoring, and asymmetric synthesis.

quant-ph↗

arXiv2025

A Backdoor-based Explainable AI Benchmark for High Fidelity Evaluation of Attributions

Attribution methods compute importance scores for input features to explain model predictions. However, assessing the faithfulness of these methods remains challenging due to the absence of attribution ground truth to model predictions. In this work, we first identify a set of fidelity criteria that reliable benchmarks for attribution methods are expected to fulfill, thereby facilitating a systematic assessment of attribution benchmarks. Next, we introduce a Backdoor-based eXplainable AI benchmark (BackX) that adheres to the desired fidelity criteria. We theoretically establish the superiority of our approach over the existing benchmarks for well-founded attribution evaluation. With extensive analysis, we further establish a standardized evaluation setup that mitigates confounding factors such as post-processing techniques and explained predictions, thereby ensuring a fair and consistent benchmarking. This setup is ultimately employed for a comprehensive comparison of existing methods using BackX. Finally, our analysis also offers insights into defending against neural Trojans by utilizing the attributions.

cs.CR↗

arXiv2025

Frequency Range Boosted Magnetometry Beyond the Spin Coherence Limit via Compressive Sensing

Free induction decay (FID) of spin precession serves as an essential tool for quantum sensing across diverse platforms. While extending spin coherence time remains critical for sensitivity enhancement, the requisite long single-shot acquisitions narrow the resolvable frequency range, establishing a fundamental ``spin coherence limit (SCL)'', according to the Nyquist Sampling Theorem. Besides, conventional spectral analysis for FID measurement suffers from frequency alias, causing signal attenuation and positional errors that compromise the measurement validity. Here, we demonstrate a general frequency-range-extended technique that overcomes SCL by leveraging compressive sensing. By applying this method to the FID magnetometer, we expand the resolvable frequency range significantly from the Nyquist-limited range of 251\,Hz to 3000\,Hz, effectively avoiding frequency alias. Our work paves the way for implementing long-coherence-time spin systems in high-sensitivity, broad-bandwidth, and alias-free magnetic field sensing.

physics.atom-ph↗

arXiv2025

Near-perfect broadband quantum memory enabled by intelligent spinwave compaction

Quantum memory, a pivotal hub in quantum information processing, is expected to achieve high-performance storage and coherent manipulation of quantum states, with memory efficiency exceeding 90% and quantum fidelity surpassing the non-cloning limit. However, the current performance falls short of these requirements due to the inherent trade-off between memory efficiency enhancement and noise amplification, which not only imposes significant demands on quantum purification but also fundamentally impedes continuous-variable quantum information processing. In this paper, we break through these constraints, enabling high-performance quantum memory and unlocking new possibilities for quantum technologies. We unveil a Hankel-transform spatiotemporal mapping for light-spinwave conversion in quantum memory, and propose an intelligent light-manipulated strategy for adaptive spinwave compaction, which can maximize the conversion efficiency and simultaneously suppress the excess noise. This strategy is experimentally demonstrated for a Raman quantum memory in warm 87Rb atomic vapor with an efficiency up to 94.6% and a low noise level of only 0.026 photons/pulse. The unconditional fidelity reaches 98.91% with an average of 1.0 photons/pulse for a 17-ns input signal. Our results successfully demonstrate a practical benchmark for broadband quantum memory, which may facilitate advancements in high-speed quantum networks, quantum state manipulation, and scalable quantum computation.

quant-ph↗

arXiv2025

Versatile, Robust, and Explosive Locomotion with Rigid and Articulated Compliant Quadrupeds

Achieving versatile and explosive motion with robustness against dynamic uncertainties is a challenging task. Introducing parallel compliance in quadrupedal design is deemed to enhance locomotion performance, which, however, makes the control task even harder. This work aims to address this challenge by proposing a general template model and establishing an efficient motion planning and control pipeline. To start, we propose a reduced-order template model-the dual-legged actuated spring-loaded inverted pendulum with trunk rotation-which explicitly models parallel compliance by decoupling spring effects from active motor actuation. With this template model, versatile acrobatic motions, such as pronking, froggy jumping, and hop-turn, are generated by a dual-layer trajectory optimization, where the singularity-free body rotation representation is taken into consideration. Integrated with a linear singularity-free tracking controller, enhanced quadrupedal locomotion is achieved. Comparisons with the existing template model reveal the improved accuracy and generalization of our model. Hardware experiments with a rigid quadruped and a newly designed compliant quadruped demonstrate that i) the template model enables generating versatile dynamic motion; ii) parallel elasticity enhances explosive motion. For example, the maximal pronking distance, hop-turn yaw angle, and froggy jumping distance increase at least by 25%, 15% and 25%, respectively; iii) parallel elasticity improves the robustness against dynamic uncertainties, including modelling errors and external disturbances. For example, the allowable support surface height variation increases by 100% for robust froggy jumping.

cs.RO↗

arXiv2025

Quantum Twin Interferometers

Quantum-correlated interferometer is a newly emerging tool in quantum technology that offers classical-limit-breaking phase sensitivity. But to date, there exists a configurational bottleneck for its practicability due to the low phase-sensitive photon numbers limited by the current detection strategies. Here we establish an innovative development termed as ``quantum twin interferometer'' with dual pairs of entangled twin beams arranged in the parallel configuration, allowing fully exploits the quantum resource through the new configuration of entangled detection. We observe the distributed phase sensing with 3 dB quantum noise reduction in phase-sensing power at the level of milliwatts, which advances the record of signal-to-noise ratio so far achieved in photon-correlated interferometers by three orders of magnitude. The developed techniques in this work can be used to revolutionize a diversity of quantum devices requiring phase measurement.

physics.atom-ph↗

arXiv2024

Quantum Locking of Intrinsic Spin Squeezed State in Earth-field-range Magnetometry

In the Earth-field range, the nonlinear Zeeman (NLZ) effect has been a bottleneck limiting the sensitivity and accuracy of atomic magnetometry from physical mechanism. To break this bottleneck, various techniques are introduced to suppress the NLZ effect. Here we revisit the spin dynamics in the Earth-field-range magnetometry and identify the existence of the intrinsic spin squeezed state (SSS) generated from the geomagnetically induced NLZ effect with the oscillating squeezing degree and squeezing axis. Such oscillating features of the SSS prevent its direct observation and as well, accessibility to magnetic sensing. To exploit quantum advantage of the intrinsic SSS in the Earth-field-range magnetometry, it's essential to lock the oscillating SSS to a persistent one. Hence, we develop a quantum locking technique to achieve a persistent SSS, benefiting from which the sensitivity of the Earth-field-range magnetometer is quantum-enhanced. This work presents an innovative way turning the drawback of NLZ effect into the quantum advantage and opens a new access to quantum-enhanced magnetometry in the Earth-field range.

physics.atom-ph↗

arXiv2024

Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density

Trustworthy machine learning necessitates meticulous regulation of model reliance on non-robust features. We propose a framework to delineate and regulate such features by attributing model predictions to the input. Within our approach, robust feature attributions exhibit a certain consistency, while non-robust feature attributions are susceptible to fluctuations. This behavior allows identification of correlation between model reliance on non-robust features and smoothness of marginal density of the input samples. Hence, we uniquely regularize the gradients of the marginal density w.r.t. the input features for robustness. We also devise an efficient implementation of our regularization to address the potential numerical instability of the underlying optimization process. Moreover, we analytically reveal that, as opposed to our marginal density smoothing, the prevalent input gradient regularization smoothens conditional or joint density of the input, which can cause limited robustness. Our experiments validate the effectiveness of the proposed method, providing clear evidence of its capability to address the feature leakage problem and mitigate spurious correlations. Extensive results further establish that our technique enables the model to exhibit robustness against perturbations in pixel values, input gradients, and density.

cs.LG↗

arXiv2024

Fast-PGM: Fast Probabilistic Graphical Model Learning and Inference

Probabilistic graphical models (PGMs) serve as a powerful framework for modeling complex systems with uncertainty and extracting valuable insights from data. However, users face challenges when applying PGMs to their problems in terms of efficiency and usability. This paper presents Fast-PGM, an efficient and open-source library for PGM learning and inference. Fast-PGM supports comprehensive tasks on PGMs, including structure and parameter learning, as well as exact and approximate inference, and enhances efficiency of the tasks through computational and memory optimizations and parallelization techniques. Concurrently, Fast-PGM furnishes developers with flexible building blocks, furnishes learners with detailed documentation, and affords non-experts user-friendly interfaces, thereby ameliorating the usability of PGMs to users across a spectrum of expertise levels. The source code of Fast-PGM is available at https://github.com/jjiantong/FastPGM.

cs.LG↗

arXiv2023

AnoOnly: Semi-Supervised Anomaly Detection with the Only Loss on Anomalies

Semi-supervised anomaly detection (SSAD) methods have demonstrated their effectiveness in enhancing unsupervised anomaly detection (UAD) by leveraging few-shot but instructive abnormal instances. However, the dominance of homogeneous normal data over anomalies biases the SSAD models against effectively perceiving anomalies. To address this issue and achieve balanced supervision between heavily imbalanced normal and abnormal data, we develop a novel framework called AnoOnly (Anomaly Only). Unlike existing SSAD methods that resort to strict loss supervision, AnoOnly suspends it and introduces a form of weak supervision for normal data. This weak supervision is instantiated through the utilization of batch normalization, which implicitly performs cluster learning on normal data. When integrated into existing SSAD methods, the proposed AnoOnly demonstrates remarkable performance enhancements across various models and datasets, achieving new state-of-the-art performance. Additionally, our AnoOnly is natively robust to label noise when suffering from data contamination. Our code is publicly available at https://github.com/cool-xuan/AnoOnly.

cs.LG↗

arXiv2023

Quantum-enhanced Electrometer based on Microwave-dressed Rydberg Atoms

Rydberg atoms have been shown remarkable performance in sensing microwave field. The sensitivity of such an electrometer based on optical readout of atomic ensemble has been demonstrated to approach the photon-shot-noise limit. However, the sensitivity can not be promoted infinitely by increasing the power of probe light due to the increased collision rates and power broadening. Compared with classical light, the use of quantum light may lead to a better sensitivity with lower number of photons. In this paper, we exploit entanglement in a microwave-dressed Rydberg electrometer to suppress the fluctuation of noise. The results show a sensitivity enhancement beating the shot noise limit in both cold and hot atom schemes. Through optimizing the transmission of optical readout, our quantum advantage can be maintained with different absorptive index of atomic vapor, which makes it possible to apply quantum light source in the absorptive electrometer.

quant-ph↗

arXiv2021

Coherence Protection of Electron Spin in Earth-field Range by All-optical Dynamic Decoupling

In recent years, unshielded atomic systems have been attracting researchers' attention, in which decoherence is one of the major problems, especially for high precision measurements. The nonlinear Zeeman effect and magnetic field gradient are the main decoherence sources of atomic electron spin in Earth-field range. Here, we propose a method to cancel out the two dominant broadening effects simultaneously by an all-optical dynamic decoupling approach based on Raman scattering in the 87Rb Zeeman sublevels. By adjusting the parameters of the Raman lasers, we realize spin control along an arbitrary direction. We analyze the state evolution of atomic spin under the Raman light control sequence in detail. The results show that both the nonlinear Zeeman effect and magnetic field gradient can be significantly suppressed.

physics.atom-ph↗