arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 433 records · Page 24Linked to original sources

arXiv2025

Dual slow-light enhanced photothermal gas spectroscopy on a silicon chip

Integrated photonic sensors have attracted significant attention recently for their potential for high-density integration. However, they face challenges in sensing gases with high sensitivity due to weak light-gas interaction. Slow light, which dramatically intensifies light-matter interaction through spatial compression of optical energy, provides a promising solution. Herein, we demonstrate a dual slow-light scheme for enhancing the sensitivity of photothermal spectroscopy (PTS) with a suspended photonic crystal waveguide (PhCW) on a CMOS-compatible silicon platform. By tailoring the dispersion of the PhCW to generate structural slow light to enhance pump absorption and probe phase modulation, we achieve a photothermal efficiency of 3.6x10-4 rad cm ppm-1 mW-1 m-1, over 1-3 orders of magnitude higher than the strip waveguides and optical fibers. With a 1-mm-long sensing PhCW incorporated in a stabilized on-chip Mach-Zehnder interferometer with a footprint of 0.6 mm2, we demonstrate acetylene detection with a sensitivity of 1.4x10-6 in terms of noise-equivalent absorption and length product (NEAL), the best among the reported photonic waveguide gas sensors to our knowledge. The dual slow-light enhanced PTS paves the way for integrated photonic gas sensors with high sensitivity, miniaturization, and cost-effective mass production.

physics.optics↗

arXiv2025

Think Consistently, Reason Efficiently: Energy-Based Calibration for Implicit Chain-of-Thought

Large Language Models (LLMs) have demonstrated strong reasoning capabilities through \emph{Chain-of-Thought} (CoT) prompting, which enables step-by-step intermediate reasoning. However, explicit CoT methods rely on discrete token-level reasoning processes that are prone to error propagation and limited by vocabulary expressiveness, often resulting in rigid and inconsistent reasoning trajectories. Recent research has explored implicit or continuous reasoning in latent spaces, allowing models to perform internal reasoning before generating explicit output. Although such approaches alleviate some limitations of discrete CoT, they generally lack explicit mechanisms to enforce consistency among reasoning steps, leading to divergent reasoning paths and unstable outcomes. To address this issue, we propose EBM-CoT, an Energy-Based Chain-of-Thought Calibration framework that refines latent thought representations through an energy-based model (EBM). Our method dynamically adjusts latent reasoning trajectories toward lower-energy, high-consistency regions in the embedding space, improving both reasoning accuracy and consistency without modifying the base language model. Extensive experiments across mathematical, commonsense, and symbolic reasoning benchmarks demonstrate that the proposed framework significantly enhances the consistency and efficiency of multi-step reasoning in LLMs.

cs.CL↗

arXiv2025

DiLA: Enhancing LLM Tool Learning with Differential Logic Layer

Considering the challenges faced by large language models (LLMs) in logical reasoning and planning, prior efforts have sought to augment LLMs with access to external solvers. While progress has been made on simple reasoning problems, solving classical constraint satisfaction problems, such as the Boolean Satisfiability Problem (SAT) and Graph Coloring Problem (GCP), remains difficult for off-the-shelf solvers due to their intricate expressions and exponential search spaces. In this paper, we propose a novel differential logic layer-aided language modeling (DiLA) approach, where logical constraints are integrated into the forward and backward passes of a network layer, to provide another option for LLM tool learning. In DiLA, LLM aims to transform the language description to logic constraints and identify initial solutions of the highest quality, while the differential logic layer focuses on iteratively refining the LLM-prompted solution. Leveraging the logic layer as a bridge, DiLA enhances the logical reasoning ability of LLMs on a range of reasoning problems encoded by Boolean variables, guaranteeing the efficiency and correctness of the solution process. We evaluate the performance of DiLA on two classic reasoning problems and empirically demonstrate its consistent outperformance against existing prompt-based and solver-aided approaches.

cs.CL↗

arXiv2025

Fast and Robust Remote Two-Qubit Gates on Distributed Qubits

Distributed quantum computing offers a potential solution to the complexity of superconducting chip hardware layouts and error correction algorithms. High-quality gates between distributed chips enable the simplification of existing error correction algorithms. This article proposes and demonstrates a remote quantum geometric gate scheme via parametric modulation. Our scheme inherits the intrinsic robustness of geometric phases. Meanwhile, by employing gradient-based optimization algorithms(Adaptive Moment Estimation) from deep learning, we design control waveforms that significantly suppress population leakage. We experimentally realize the rapid remote SWAP and $\sqrt{\text{SWAP}}$ gates with high fidelity, completing operation in about 30 ns. The gate error of SWAP ($\sqrt{\text{SWAP}}$) is 1.16\% (0.91\%) after excluding the effect of energy relaxation. The simulation demonstrate that this scheme can be implemented in the distributed chips connected by cables extending several meters. Our results highlight the effectiveness of the proposed protocol in enabling modular quantum processors, offering a promising path toward the realization of fault-tolerant quantum computation.

quant-ph↗

arXiv2025

LLM-Assisted Tool for Joint Generation of Formulas and Functions in Rule-Based Verification of Map Transformations

High-definition map transformations are essential in autonomous driving systems, enabling interoperability across tools. Ensuring their semantic correctness is challenging, since existing rule-based frameworks rely on manually written formulas and domain-specific functions, limiting scalability. In this paper, We present an LLM-assisted pipeline that jointly generates logical formulas and corresponding executable predicates within a computational FOL framework, extending the map verifier in CommonRoad scenario designer with elevation support. The pipeline leverages prompt-based LLM generation to produce grammar-compliant rules and predicates that integrate directly into the existing system. We implemented a prototype and evaluated it on synthetic bridge and slope scenarios. The results indicate reduced manual engineering effort while preserving correctness, demonstrating the feasibility of a scalable, semi-automated human-in-the-loop approach to map-transformation verification.

cs.SE↗

arXiv2025

Kimi Linear: An Expressive, Efficient Attention Architecture

We introduce Kimi Linear, a hybrid linear attention architecture that, for the first time, outperforms full attention under fair comparisons across various scenarios -- including short-context, long-context, and reinforcement learning (RL) scaling regimes. At its core lies Kimi Delta Attention (KDA), an expressive linear attention module that extends Gated DeltaNet with a finer-grained gating mechanism, enabling more effective use of limited finite-state RNN memory. Our bespoke chunkwise algorithm achieves high hardware efficiency through a specialized variant of the Diagonal-Plus-Low-Rank (DPLR) transition matrices, which substantially reduces computation compared to the general DPLR formulation while remaining more consistent with the classical delta rule. We pretrain a Kimi Linear model with 3B activated parameters and 48B total parameters, based on a layerwise hybrid of KDA and Multi-Head Latent Attention (MLA). Our experiments show that with an identical training recipe, Kimi Linear outperforms full MLA with a sizeable margin across all evaluated tasks, while reducing KV cache usage by up to 75% and achieving up to 6 times decoding throughput for a 1M context. These results demonstrate that Kimi Linear can be a drop-in replacement for full attention architectures with superior performance and efficiency, including tasks with longer input and output lengths. To support further research, we open-source the KDA kernel and vLLM implementations, and release the pre-trained and instruction-tuned model checkpoints.

cs.CL↗

arXiv2025

Static for Dynamic: Towards a Deeper Understanding of Dynamic Facial Expressions Using Static Expression Data

Dynamic facial expression recognition (DFER) infers emotions from the temporal evolution of expressions, unlike static facial expression recognition (SFER), which relies solely on a single snapshot. This temporal analysis provides richer information and promises greater recognition capability. However, current DFER methods often exhibit unsatisfied performance largely due to fewer training samples compared to SFER. Given the inherent correlation between static and dynamic expressions, we hypothesize that leveraging the abundant SFER data can enhance DFER. To this end, we propose Static-for-Dynamic (S4D), a unified dual-modal learning framework that integrates SFER data as a complementary resource for DFER. Specifically, S4D employs dual-modal self-supervised pre-training on facial images and videos using a shared Vision Transformer (ViT) encoder-decoder architecture, yielding improved spatiotemporal representations. The pre-trained encoder is then fine-tuned on static and dynamic expression datasets in a multi-task learning setup to facilitate emotional information interaction. Unfortunately, vanilla multi-task learning in our study results in negative transfer. To address this, we propose an innovative Mixture of Adapter Experts (MoAE) module that facilitates task-specific knowledge acquisition while effectively extracting shared knowledge from both static and dynamic expression data. Extensive experiments demonstrate that S4D achieves a deeper understanding of DFER, setting new state-of-the-art performance on FERV39K, MAFW, and DFEW benchmarks, with weighted average recall (WAR) of 53.65\%, 58.44\%, and 76.68\%, respectively. Additionally, a systematic correlation analysis between SFER and DFER tasks is presented, which further elucidates the potential benefits of leveraging SFER.

cs.CV↗

arXiv2025

Probing dark-state electromagnetic form factors at future linear electron-positron colliders with polarized beams

We study the electromagnetic form factors of electrically neutral dark-sector particles in this paper. Concretely speaking, we focus on operators that can lead to interactions of fermionic dark states with the Standard-Model (SM) photon, including both dimension-5 (the magnetic and electric dipole moments) and dimension-6 (the anapole moment and charge radius) operators. Correspondingly, instead of the SM-photon form factors, we employ hypercharge gauge-field form factors that can induce additional couplings to the SM $Z$-boson. Utilizing the mono-photon production channel at future linear electron-positron colliders, the International Linear Collider (ILC) and the Compact Linear Collider (CLIC), we demonstrate that tuning beam polarization allows for simultaneous enhancement of signal-event rates and suppression of dominant SM background events, i.e.~neutrino-induced processes, and hence improvement of sensitivity reach. Our analysis reveals that ILC and CLIC can probe electromagnetic form factors of dark-sector particles up to about two orders of magnitude beyond existing experimental limits. We also estimate, at the order-of-magnitude level, the validity range of the effective-field-theory (EFT) approach we adopt, finding that for the dimension-5 operators the EFT remains valid in most of the parameter regions to which the ILC and CLIC can be sensitive, while it breaks down in the entire sensitivity region for the dimension-6 operators.

hep-ph↗

arXiv2025

DARTS: A Drone-Based AI-Powered Real-Time Traffic Incident Detection System

Rapid and reliable incident detection is critical for reducing crash-related fatalities, injuries, and congestion. However, conventional methods, such as closed-circuit television, dashcam footage, and sensor-based detection, separate detection from verification, suffer from limited flexibility, and require dense infrastructure or high penetration rates, restricting adaptability and scalability to shifting incident hotspots. To overcome these challenges, we developed DARTS, a drone-based, AI-powered real-time traffic incident detection system. DARTS integrates drones' high mobility and aerial perspective for adaptive surveillance, thermal imaging for better low-visibility performance and privacy protection, and a lightweight deep learning framework for real-time vehicle trajectory extraction and incident detection. The system achieved 99% detection accuracy on a self-collected dataset and supports simultaneous online visual verification, severity assessment, and incident-induced congestion propagation monitoring via a web-based interface. In a field test on Interstate 75 in Florida, DARTS detected and verified a rear-end collision 12 minutes earlier than the local transportation management center and monitored incident-induced congestion propagation, suggesting potential to support faster emergency response and enable proactive traffic control to reduce congestion and secondary crash risk. Crucially, DARTS's flexible deployment architecture reduces dependence on frequent physical patrols, indicating potential scalability and cost-effectiveness for use in remote areas and resource-constrained settings. This study presents a promising step toward a more flexible and integrated real-time traffic incident detection system, with significant implications for the operational efficiency and responsiveness of modern transportation management.

cs.RO↗

arXiv2025

Mining Large Independent Sets on Massive Graphs

The Maximum Independent Set problem is fundamental for extracting conflict-free structure from large graphs, with applications in scheduling, recommendation, and network analysis. However, existing heuristics can stagnate when search schedules are fixed and information from past solutions is underused, leading to wasted effort in low-quality regions of the search space. We present ARCIS, an efficient algorithm for mining large independent sets on massive graphs. ARCIS couples two main components. The first is an adaptive restart policy that refreshes exploration when progress slows. The second is Consensus-Guided Vertex Fixing, which restricts the search to the non-consensus region of the graph by fixing vertices consistently observed within a round. The consensus is maintained as a running intersection within each round, and because it is recomputed at every restart, the fixing is reversible. Vertices that later lose support are automatically unfixed and their neighborhoods re-enter the working graph, which corrects occasional mistakes while preserving progress. Experiments on 222 graphs from four benchmark suites show that ARCIS attains the best or tied-best solution quality in most instances while delivering competitive runtime and low variability. Ablation studies isolate the impact of each component, indicating that ARCIS is a practical and robust method for large-scale graph mining.

cs.AI↗

arXiv2025

MoPFormer: Motion-Primitive Transformer for Wearable-Sensor Activity Recognition

Human Activity Recognition (HAR) with wearable sensors is challenged by limited interpretability, which significantly impacts cross-dataset generalization. To address this challenge, we propose Motion-Primitive Transformer (MoPFormer), a novel self-supervised framework that enhances interpretability by tokenizing inertial measurement unit signals into semantically meaningful motion primitives and leverages a Transformer architecture to learn rich temporal representations. MoPFormer comprises two stages. The first stage is to partition multi-channel sensor streams into short segments and quantize them into discrete ``motion primitive'' codewords, while the second stage enriches those tokenized sequences through a context-aware embedding module and then processes them with a Transformer encoder. The proposed MoPFormer can be pre-trained using a masked motion-modeling objective that reconstructs missing primitives, enabling it to develop robust representations across diverse sensor configurations. Experiments on six HAR benchmarks demonstrate that MoPFormer not only outperforms state-of-the-art methods but also successfully generalizes across multiple datasets. More importantly, the learned motion primitives significantly enhance both interpretability and cross-dataset performance by capturing fundamental movement patterns that remain consistent across similar activities, regardless of dataset origin.

cs.CV↗

arXiv2025

SN 2024iss: A Double-peaked Type IIb Supernova with Evidence of Circumstellar Interaction

We present optical, ultraviolet, and X-ray observations of supernova (SN) 2024iss, a Type IIb SN that shows a prominent double-peaked light curve. We modeled the first peak with a semianalytical shock-cooling model and the X-ray emission with a free-free model. We compare the envelope radius and mass-loss rate with other Type IIb SNe to explore the relationships between the progenitor envelope and the circumstellar material (CSM). The shock-cooling peak in the $V$-band light curve reached $M_V = -17.33\pm 0.26$mag, while the $^{56}$Ni-powered second peak attained $M_V = -17.43\pm 0.26$mag. Early spectra show an photospheric velocity of $\sim19,400\,km\,s^{-1}$ at 3.82days from the H$α$ P~Cygni profile. The Balmer lines persist at least +87 days after the explosion, characterizing hydrogen-rich ejecta. Modeling the first light-curve peak suggests an extended envelope with a mass of $0.11\pm0.04\,M_{\odot}$ and a radius of $244\pm43~R_{\odot}$. Fitting the second light-curve peak with an Arnett-like model indicates a typical $^{56}$Ni mass of $ 0.117\pm0.013~M_{\odot}$ and a relatively low ejecta mass of $1.272\pm0.343\,M_{\odot}$. X-ray observations reveal bright thermal bremsstrahlung emission and indicate a mass-loss rate of $1.6\times10^{-5}\ M_{\odot} \ \rm{yr}^{-1}$. SN 2024iss occupies a transitional position between the two subclasses of extended (eIIb) and compact (cIIb) Type IIb SNe. Its envelope radius and pre-explosion mass-loss rate appear to be correlated as theoretically predicted. The observational properties of SN 2024iss are compatible with a binary interaction scenario being the dominant mechanism for envelope stripping. Furthermore, the low column density of neutral hydrogen suggests a compact CSM with an outer radius of $\lesssim1.3\times10^{14}$ cm, indicating that the progenitor star experienced eruptive mass loss within $\sim4\,yr$ of its terminal explosion.

astro-ph.HE↗

arXiv2025

ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibility Data

The advent of single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) offers an innovative perspective for deciphering regulatory mechanisms by assembling a vast repository of single-cell chromatin accessibility data. While foundation models have achieved significant success in single-cell transcriptomics, there is currently no foundation model for scATAC-seq that supports zero-shot high-quality cell identification and comprehensive multi-omics analysis simultaneously. Key challenges lie in the high dimensionality and sparsity of scATAC-seq data, as well as the lack of a standardized schema for representing open chromatin regions (OCRs). Here, we present ChromFound, a foundation model tailored for scATAC-seq. ChromFound utilizes a hybrid architecture and genome-aware tokenization to effectively capture genome-wide long contexts and regulatory signals from dynamic chromatin landscapes. Pretrained on 1.97 million cells from 30 tissues and 6 disease conditions, ChromFound demonstrates broad applicability across 6 diverse tasks. Notably, it achieves robust zero-shot performance in generating universal cell representations and exhibits excellent transferability in cell type annotation and cross-omics prediction. By uncovering enhancer-gene links undetected by existing computational methods, ChromFound offers a promising framework for understanding disease risk variants in the noncoding genome.

q-bio.GN↗

arXiv2025

Harnessing Intrinsic Noise for Quantum Simulation of Open Quantum Systems

Simulating open quantum systems on quantum computers presents a fundamental challenge: open quantum dynamics are intrinsically nonunitary, whereas quantum computers operate through unitary evolution. Conventional approaches overcome this mismatch by encoding nonunitary processes into unitary circuits, but such methods incur substantial overhead in both qubits and gates. Here, we propose an alternative perspective. Quantum processors are themselves open systems, inherently subject to noise. Instead of correcting all errors and then encoding nonunitary dynamics with unitary logical qubits and gates, we show how noise can be harnessed as a computational resource. We develop a noise-assisted quantum algorithm that selectively preserves physical noise to emulate nonunitary channels, enabling efficient simulation of open quantum dynamics with minimal qubit requirements. Our approach applies both to noisy intermediate-scale quantum (NISQ) devices and future fault-tolerant architectures. By leveraging intrinsic noise, this method circumvents the need to encode nonunitary dynamics into unitary gates and relaxes fidelity requirements on physical qubits, thereby reducing the overhead of quantum error correction. This framework reframes noise from a limitation into a resource, opening new directions for practical quantum simulation of open systems

quant-ph↗

arXiv2025

PLAN: Proactive Low-Rank Allocation for Continual Learning

Continual learning (CL) requires models to continuously adapt to new tasks without forgetting past knowledge. In this work, we propose \underline{P}roactive \underline{L}ow-rank \underline{A}llocatio\underline{N} (PLAN), a framework that extends Low-Rank Adaptation (LoRA) to enable efficient and interference-aware fine-tuning of large pre-trained models in CL settings. PLAN proactively manages the allocation of task-specific subspaces by introducing orthogonal basis vectors for each task and optimizing them through a perturbation-based strategy that minimizes conflicts with previously learned parameters. Furthermore, PLAN incorporates a novel selection mechanism that identifies and assigns basis vectors with minimal sensitivity to interference, reducing the risk of degrading past knowledge while maintaining efficient adaptation to new tasks. Empirical results on standard CL benchmarks demonstrate that PLAN consistently outperforms existing methods, establishing a new state-of-the-art for continual learning with foundation models.

cs.LG↗

arXiv2025

Domain wall induced topological Hall effect in the chiral-lattice ferromagnet Fe$_x$TaS$_2$

Magnetic topology and its associated emergent phenomena are central to realizing intriguing quantum states and spintronics functionalities. Designing spin textures to achieve strong and distinct electrical responses remains a significant challenge. Layered transition metal dichalcogenides offer a versatile platform for tailoring structural and magnetic properties, enabling access to a wide spectrum of topological magnetic states. Here, we report a domain-wall-driven, large, and tunable topological Hall effect (THE) in a non-centrosymmetric intercalated transition metal dichalcogenides series Fe$_x$TaS$_2$. By systematically varying the Fe intercalation level, we exert precise control over the magnetic ground states, allowing manipulation of the topological Hall effect. Real-space magnetic force microscopy (MFM) provides direct evidence of periodic magnetic stripe domain formation, confirming the microscopic origin of the observed topological transport phenomena. Our findings establish a promising way for tuning the topology of domains to generate substantial electromagnetic responses in layered magnetic materials.

cond-mat.mtrl-sci↗

arXiv2025

EmbodiedBrain: Expanding Performance Boundaries of Task Planning for Embodied Intelligence

The realization of Artificial General Intelligence (AGI) necessitates Embodied AI agents capable of robust spatial perception, effective task planning, and adaptive execution in physical environments. However, current large language models (LLMs) and multimodal LLMs (MLLMs) for embodied tasks suffer from key limitations, including a significant gap between model design and agent requirements, an unavoidable trade-off between real-time latency and performance, and the use of unauthentic, offline evaluation metrics. To address these challenges, we propose EmbodiedBrain, a novel vision-language foundation model available in both 7B and 32B parameter sizes. Our framework features an agent-aligned data structure and employs a powerful training methodology that integrates large-scale Supervised Fine-Tuning (SFT) with Step-Augumented Group Relative Policy Optimization (Step-GRPO), which boosts long-horizon task success by integrating preceding steps as Guided Precursors. Furthermore, we incorporate a comprehensive reward system, including a Generative Reward Model (GRM) accelerated at the infrastructure level, to improve training efficiency. For enable thorough validation, we establish a three-part evaluation system encompassing General, Planning, and End-to-End Simulation Benchmarks, highlighted by the proposal and open-sourcing of a novel, challenging simulation environment. Experimental results demonstrate that EmbodiedBrain achieves superior performance across all metrics, establishing a new state-of-the-art for embodied foundation models. Towards paving the way for the next generation of generalist embodied agents, we open-source all of our data, model weight, and evaluating methods, which are available at https://zterobot.github.io/EmbodiedBrain.github.io.

cs.CV↗

arXiv2025

Provably Efficient Reward Transfer in Reinforcement Learning with Discrete Markov Decision Processes

In this paper, we propose a new solution to reward adaptation (RA) in reinforcement learning, where the agent adapts to a target reward function based on one or more existing source behaviors learned a priori under the same domain dynamics but different reward functions. While learning the target behavior from scratch is possible, it is often inefficient given the available source behaviors. Our work introduces a new approach to RA through the manipulation of Q-functions. Assuming the target reward function is a known function of the source reward functions, we compute bounds on the Q-function and present an iterative process (akin to value iteration) to tighten these bounds. Such bounds enable action pruning in the target domain before learning even starts. We refer to this method as "Q-Manipulation" (Q-M). The iteration process assumes access to a lite-model, which is easy to provide or learn. We formally prove that Q-M, under discrete domains, does not affect the optimality of the returned policy and show that it is provably efficient in terms of sample complexity in a probabilistic sense. Q-M is evaluated in a variety of synthetic and simulation domains to demonstrate its effectiveness, generalizability, and practicality.

cs.LG↗