arXiv subjects

Peiyi Li

Publications and source records attributed to Peiyi Li.

16 recordsLinked to original sources

arXiv2026

Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillation

Sampled-token on-policy distillation (OPD) efficiently transfers capabilities from teacher to student using student-generated tokens, requiring teacher probabilities only for sampled tokens. Yet it frequently suffers from diversity distillation failure: the student's pass@1 improves while its pass@$k$ plateaus, failing to inherit the teacher's diversity. To explain this, we introduce First-Order Local Entropy Influence, a signed first-order proxy that decouples each update's entropy effect into the teacher--student log-probability gap and the student's local probability structure, and empirically links entropy contraction to negative-influence positions. Motivated by this, we propose Influence-Directed Adaptive On-Policy Distillation (IDA-OPD): rather than relying on costly full-vocabulary Forward-KL objectives, it preserves entropy-expanding updates while replacing entropy-contracting ones with divergence-adaptive advantage shrinkage, using only the teacher's sampled-token log-probability. Experiments on reasoning-oriented distillation show IDA-OPD consistently improves pass@$k$, inheriting the teacher's diversity through distillation, matches the strongest teacher-informed methods at strictly lower cost, and broadly maintains vanilla OPD's pass@1, all without full-vocabulary teacher information.

cs.CL↗

arXiv2026

MemeBench: What LVLMs Miss When Interpreting Culture-Dependent Memes

Large vision-language models have improved at describing visual content, but accurate descriptions do not ensure interpretation when meaning depends on knowledge beyond the pixels. Memes expose this gap because they rely on cultural entities, background knowledge, and community conventions. Most meme benchmarks reduce interpretation to labels or holistic scores, obscuring where an explanation breaks down. We introduce MemeBench, a diagnostic benchmark of 1,253 Chinese and English memes with human-written references and quality-controlled VIKR annotations, centered on anime, comics, games, and adjacent online subcultures. Its VIKR schema decomposes explanations into Visual clues, Identity links, Knowledge units, and Reasoning mechanisms. Across 26 LVLMs, every model covers visible content more reliably than the knowledge needed to interpret it, and even the strongest retains a 22.6% Visual-Knowledge gap. To test whether this diagnosis can guide improvement, we introduce KAR, an entity-guided retrieval baseline built on CultureBase. Across four controlled models, KAR raises VIKR Success by 3.6-7.4% and, compared with generic retrieval, repairs more answers and breaks fewer. Yet both retrieval conditions improve Identity and Knowledge while reducing Visual coverage in every comparison. MemeBench reveals whether an interpretation succeeds, what is missing, and whether targeted evidence fills the diagnosed gap.

cs.AI↗

arXiv2026

TextFlux: An OCR-Free DiT Model for High-Fidelity Multilingual Scene Text Synthesis

Diffusion-based scene text synthesis has progressed rapidly, yet existing methods commonly rely on additional visual conditioning modules and require large-scale annotated data to support multilingual generation. In this work, we revisit the necessity of complex auxiliary modules and further explore an approach that simultaneously ensures glyph accuracy and achieves high-fidelity scene integration, by leveraging diffusion models' inherent capabilities for contextual reasoning. To this end, we introduce TextFlux, a DiT-based framework that enables multilingual scene text synthesis. The advantages of TextFlux can be summarized as follows: (1) OCR-free model architecture. TextFlux eliminates the need for OCR encoders (additional visual conditioning modules) that are specifically used to extract visual text-related features. (2) Strong multilingual scalability. TextFlux is effective in low-resource multilingual settings, and achieves strong performance in newly added languages with fewer than 1,000 samples. (3) Streamlined training setup. TextFlux is trained with only 1% of the training data required by competing methods. (4) Controllable multi-line text generation. TextFlux offers flexible multi-line synthesis with precise line-level control, outperforming methods restricted to single-line or rigid layouts. Extensive experiments and visualizations demonstrate that TextFlux outperforms previous methods in both qualitative and quantitative evaluations.

cs.CV↗

arXiv2026

Efficient Causal Structure Learning via Modular Subgraph Integration

Learning causal structures from observational data remains a fundamental yet computationally intensive task, particularly in high-dimensional settings where existing methods face challenges such as the super-exponential growth of the search space and increasing computational demands. To address this, we introduce VISTA (Voting-based Integration of Subgraph Topologies for Acyclicity), a modular framework that decomposes the global causal structure learning problem into local subgraphs based on Markov Blankets. The global integration is achieved through a weighted voting mechanism that penalizes low-support edges via exponential decay, filters unreliable ones with an adaptive threshold, and ensures acyclicity using a Feedback Arc Set (FAS) algorithm. The framework is model-agnostic, imposing no assumptions on the inductive biases of base learners, is compatible with arbitrary data settings without requiring specific structural forms, and fully supports parallelization. We also theoretically establish finite-sample error bounds for VISTA, and prove its asymptotic consistency under mild conditions. Extensive experiments on both synthetic and real datasets consistently demonstrate the effectiveness of VISTA, yielding notable improvements in both accuracy and efficiency over a wide range of base learners.

stat.ML↗

arXiv2025

Improving VQA Reliability: A Dual-Assessment Approach with Self-Reflection and Cross-Model Verification

Vision-language models (VLMs) have demonstrated significant potential in Visual Question Answering (VQA). However, the susceptibility of VLMs to hallucinations can lead to overconfident yet incorrect answers, severely undermining answer reliability. To address this, we propose Dual-Assessment for VLM Reliability (DAVR), a novel framework that integrates Self-Reflection and Cross-Model Verification for comprehensive uncertainty estimation. The DAVR framework features a dual-pathway architecture: one pathway leverages dual selector modules to assess response reliability by fusing VLM latent features with QA embeddings, while the other deploys external reference models for factual cross-checking to mitigate hallucinations. Evaluated in the Reliable VQA Challenge at ICCV-CLVL 2025, DAVR achieves a leading $Φ_{100}$ score of 39.64 and a 100-AUC of 97.22, securing first place and demonstrating its effectiveness in enhancing the trustworthiness of VLM responses.

cs.CV↗

arXiv2025

A Quantum Network Processor Unit for Distributed Quantum Computing

As quantum computing progresses, the need for scalable solutions to address large-scale computational problems has become critical. Quantum supercomputers are the next upcoming frontier by enabling multiple quantum processors to collaborate effectively to solve large-scale computational problems. The emergence of quantum supercomputers necessitates an efficient interface to manage the quantum communication protocols between quantum processors. In this paper, we propose the Quantum Network Processing Unit (QNPU), which enables quantum applications to efficiently scale beyond the capacity of individual quantum processors, serving as a critical building block for future quantum supercomputers. The QNPU works alongside the Quantum Processing Unit (QPU) in our decoupled processing units architecture, where the QPU handles local quantum operations while the QNPU manages quantum communication between nodes. We design a comprehensive instruction set architecture (ISA) for the QNPU with high-level communication protocol abstractions, implemented via micro-operations that manage EPR resources, quantum operations, and classical communication. To facilitate programming, we introduce DistQASM, which extends OpenQASM with distributed quantum operations. We then propose a microarchitecture featuring both scalar and superscalar QNPU designs to enhance performance for communication-intensive quantum workloads. Finally, we evaluate the performance of our proposed QNPU design with distributed quantum workloads and demonstrate that the QNPU significantly improves the efficiency of communication between quantum nodes, paving the way for quantum supercomputing.

quant-ph↗

arXiv2025

Diving into Mitigating Hallucinations from a Vision Perspective for Large Vision-Language Models

Object hallucination in Large Vision-Language Models (LVLMs) significantly impedes their real-world applicability. As the primary component for accurately interpreting visual information, the choice of visual encoder is pivotal. We hypothesize that the diverse training paradigms employed by different visual encoders instill them with distinct inductive biases, which leads to their diverse hallucination performances. Existing benchmarks typically focus on coarse-grained hallucination detection and fail to capture the diverse hallucinations elaborated in our hypothesis. To systematically analyze these effects, we introduce VHBench-10, a comprehensive benchmark with approximately 10,000 samples for evaluating LVLMs across ten fine-grained hallucination categories. Our evaluations confirm encoders exhibit unique hallucination characteristics. Building on these insights and the suboptimality of simple feature fusion, we propose VisionWeaver, a novel Context-Aware Routing Network. It employs global visual features to generate routing signals, dynamically aggregating visual features from multiple specialized experts. Comprehensive experiments confirm the effectiveness of VisionWeaver in significantly reducing hallucinations and improving overall model performance.

cs.CV↗

arXiv2025

Online Item Cold-Start Recommendation with Popularity-Aware Meta-Learning

With the rise of e-commerce and short videos, online recommender systems that can capture users' interests and update new items in real-time play an increasingly important role. In both online and offline recommendation systems, the cold-start problem caused by interaction sparsity has been impacting the effectiveness of recommendations for cold-start items. Many cold-start scheme based on fine-tuning or knowledge transferring shows excellent performance on offline recommendation. Yet, these schemes are infeasible for online recommendation on streaming data pipelines due to different training method, computational overhead and time constraints. Inspired by the above questions, we propose a model-agnostic recommendation algorithm called Popularity-Aware Meta-learning (PAM), to address the item cold-start problem under streaming data settings. PAM divides the incoming data into different meta-learning tasks by predefined item popularity thresholds. The model can distinguish and reweight behavior-related and content-related features in each task based on their different roles in different popularity levels, thus adapting to recommendations for cold-start samples. These task-fixing design significantly reduces additional computation and storage costs compared to offline methods. Furthermore, PAM also introduced data augmentation and an additional self-supervised loss specifically designed for low-popularity tasks, leveraging insights from high-popularity samples. This approach effectively mitigates the issue of inadequate supervision due to the scarcity of cold-start samples. Experimental results across multiple public datasets demonstrate the superiority of our approach over other baseline methods in addressing cold-start challenges in online streaming data scenarios.

cs.IR↗

arXiv2025

USM: Unbiased Survey Modeling for Limiting Negative User Experiences in Recommendation Systems

Reducing negative user experiences is essential for the success of recommendation platforms. Exposing users to inappropriate content could not only adversely affect users' psychological well-beings, but also potentially drive users away from the platform, sabotaging the platform's long-term success. However, recommendation algorithms tend to weigh more heavily on positive feedback signals due to the scarcity of negative ones, which may result in the neglect of valuable negative user feedback. In this paper, we propose an approach aimed at limiting negative user experiences. Our method primarily relies on distributing in-feed surveys to the users, modeling the users' feedback collected from the survey, and integrating the model predictions into the recommendation system. We further enhance the baseline survey model by integrating the Learning Hidden Unit Contributions module and the Squeeze-and-Excitation module. In addition, we strive to resolve the problem of response Bias by applying a survey-submit model; The A/B testing results indicate a reduction in survey sexual rate and survey inappropriate rate, ranging from -1.44\% to -3.9\%. Additionally, we compared our methods against an online baseline that does not incorporate our approach. The results indicate that our approach significantly reduces the report rate and dislike rate by 1\% to 2.27\% compared to the baseline, confirming the effectiveness of our methods in enhancing user experience. After we launched the survey model based our approach on our platform, the model is able to bring reductions of 1.75\%, 2.57\%, 2.06\% on reports, dislikes, survey inappropriate rate, respectively.

cs.IR↗

arXiv2024

QuTracer: Mitigating Quantum Gate and Measurement Errors by Tracing Subsets of Qubits

Quantum error mitigation plays a crucial role in the current noisy-intermediate-scale-quantum (NISQ) era. As we advance towards achieving a practical quantum advantage in the near term, error mitigation emerges as an indispensable component. One notable prior work, Jigsaw, demonstrates that measurement crosstalk errors can be effectively mitigated by measuring subsets of qubits. Jigsaw operates by running multiple copies of the original circuit, each time measuring only a subset of qubits. The localized distributions yielded from measurement subsetting suffer from less crosstalk and are then used to update the global distribution, thereby achieving improved output fidelity. Inspired by the idea of measurement subsetting, we propose QuTracer, a framework designed to mitigate both gate and measurement errors in subsets of qubits by tracing the states of qubit subsets throughout the computational process. In order to achieve this goal, we introduce a technique, qubit subsetting Pauli checks (QSPC), which utilizes circuit cutting and Pauli Check Sandwiching (PCS) to trace the qubit subsets distribution to mitigate errors. The QuTracer framework can be applied to various algorithms including, but not limited to, VQE, QAOA, quantum arithmetic circuits, QPE, and Hamiltonian simulations. In our experiments, we perform both noisy simulations and real device experiments to demonstrate that QuTracer is scalable and significantly outperforms the state-of-the-art approaches.

quant-ph↗

arXiv2023

Enhancing Virtual Distillation with Circuit Cutting for Quantum Error Mitigation

Virtual distillation is a technique that aims to mitigate errors in noisy quantum computers. It works by preparing multiple copies of a noisy quantum state, bridging them through a circuit, and conducting measurements. As the number of copies increases, this process allows for the estimation of the expectation value with respect to a state that approaches the ideal pure state rapidly. However, virtual distillation faces a challenge in realistic scenarios: preparing multiple copies of a quantum state and bridging them through a circuit in a noisy quantum computer will significantly increase the circuit size and introduce excessive noise, which will degrade the performance of virtual distillation. To overcome this challenge, we propose an error mitigation strategy that uses circuit-cutting technology to cut the entire circuit into fragments. With this approach, the fragments responsible for generating the noisy quantum state can be executed on a noisy quantum device, while the remaining fragments are efficiently simulated on a noiseless classical simulator. By running each fragment circuit separately on quantum and classical devices and recombining their results, we can reduce the noise accumulation and enhance the effectiveness of the virtual distillation technique. Our strategy has good scalability in terms of both runtime and computational resources. We demonstrate our strategy's effectiveness through noisy simulation and experiments on a real quantum device.

quant-ph↗

arXiv2023

Transformation of hexagonal Lu to cubic LuH$_{2+x}$ single-crystalline films

With the recent report of near ambient superconductivity at room temperature in the N-doped lutetium hydride (Lu-H-N) system, the understanding of cubic Lu-H compounds has attracted worldwide attention. Generally, compared to polycrystal structures with non-negligible impurities, the single-crystalline form of materials with high purity can provide an opportunity to show their hidden properties. However, the experimental synthesis of single-crystalline cubic Lu-H compounds has not been reported thus far. Here, we developed an easy way to synthesize highly pure LuH$_{2+x}$ single-crystalline films by the post-annealing of Lu single-crystalline films (purity of 99.99%) in the H$_2$ atmosphere. The crystal and electronic structures of films were characterized by X-ray diffraction, Raman spectroscopy, and electrical transport. Interestingly, Lu films are silver-white and metallic, whereas their transformed LuH$_{2+x}$ films become purple-red and insulating, indicating the formation of an unreported electronic state of Lu-H compounds possibly. Our work provides a novel route to synthesize and explore more single-crystalline Lu-H compounds.

cond-mat.mtrl-sci↗

arXiv2023

Synthesis of single-crystalline LuN films

In the nitrogen-doped lutetium hydride (Lu-H-N) system, the presence of Lu-N chemical bonds plays a key role in the emergence of possible room-temperature superconductivity at near ambient pressure. However, due to the synthesis of single-crystalline LuN being a big challenge, the understanding of LuN is insufficient thus far. Here, we report on the epitaxial growth of single-crystalline LuN films. The crystal structures of LuN films were characterized by high-resolution X-ray diffraction. The measurement of low-temperature electrical transport indicates the LuN film is semiconducting from 300 to 2 K, yielding an activation gap of $\sim$ 0.02 eV. Interestingly, negative magnetoresistances can be observed below 12 K, which can result from the defects and magnetic impurities in LuN films. Our results uncover the electronic and magnetic properties of single-crystalline LuN films.

cond-mat.mtrl-sci↗

arXiv2023

Electronic and magnetic properties of Lu and LuH$_2$

Clarifying the electronic and magnetic properties of lutetium, lutetium dihydride, and lutetium oxide is very helpful to understand the emergent phenomena in lutetium-based compounds (such as room-temperature superconductivity). However, this kind of study is still scarce at present. Here, we report on the electronic and magnetic properties of lutetium metals, lutetium dihydride powders, and lutetium oxide powders. Crystal structures and chemical compositions of these samples were characterized by X-ray diffraction and X-ray photoemission spectroscopy, respectively. Electrical transport measurements show that the resistance of lutetium has a linear behavior depending on temperature, whereas the resistance of lutetium dihydride powders is independent of temperature. More interestingly, paramagnetism-ferromagnetism-spin glass transitions were observed at near 240 and 200 K, respectively, in lutetium metals. Our work uncovered the complex magnetic properties of Lu-based compounds.

cond-mat.mtrl-sci↗

arXiv2023

Folding-Free ZNE: A Comprehensive Quantum Zero-Noise Extrapolation Approach for Mitigating Depolarizing and Decoherence Noise

Quantum computers in the NISQ era are prone to noise. A range of quantum error mitigation techniques has been proposed to address this issue. Zero-noise extrapolation (ZNE) stands out as a promising one. ZNE involves increasing the noise levels in a circuit and then using extrapolation to infer the zero noise case from the noisy results obtained. This paper presents a novel ZNE approach that does not require circuit folding or noise scaling to mitigate depolarizing and/or decoherence noise. To mitigate depolarizing noise, we propose leveraging the extreme/infinite noisy case, which allows us to avoid circuit folding. Specifically, the circuit output with extreme noise becomes the maximally mixed state. We show that using circuit-reliability metrics, simple linear extrapolation can effectively mitigate depolarizing noise. With decoherence noise, different states decay into the all-zero state at a rate that depends on the number of excited states and time. Therefore, we propose a state- and latency-aware exponential extrapolation that does not involve folding or scaling. When dealing with a quantum system affected by both decoherence and depolarizing noise, we propose to use our two mitigation techniques in sequence: first applying decoherence error mitigation, followed by depolarizing error mitigation. A common limitation of ZNE schemes is that if the circuit of interest suffers from high noise, scaling-up noise levels could not provide useful data for extrapolation. We propose using circuit-cut techniques to break a large quantum circuit into smaller sub-circuits to overcome this limitation. This way, the noise levels of the sub-circuits are lower than the original circuit, and ZNE can become more effective in mitigating their noises.

quant-ph↗

arXiv2022

Not All SWAPs Have the Same Cost: A Case for Optimization-Aware Qubit Routing

Despite rapid advances in quantum computing technologies, the qubit connectivity limitation remains to be a critical challenge. Both near-term NISQ quantum computers and relatively long-term scalable quantum architectures do not offer full connectivity. As a result, quantum circuits may not be directly executed on quantum hardware, and a quantum compiler needs to perform qubit routing to make the circuit compatible with the device layout. During the qubit routing step, the compiler inserts SWAP gates and performs circuit transformations. Given the connectivity topology of the target hardware, there are typically multiple qubit routing candidates. The state-of-the-art compilers use a cost function to evaluate the number of SWAP gates for different routes and then select the one with the minimum number of SWAP gates. After qubit routing, the quantum compiler performs gate optimizations upon the circuit with the newly inserted SWAP gates. In this paper, we observe that the aforementioned qubit routing is not optimal, and qubit routing should \textit{not} be independent on subsequent gate optimizations. We find that with the consideration of gate optimizations, not all of the SWAP gates have the same basis-gate cost. These insights lead to the development of our qubit routing algorithm, NASSC (Not All Swaps have the Same Cost). NASSC is the first algorithm that considers the subsequent optimizations during the routing step. Our optimization-aware qubit routing leads to better routing decisions and benefits subsequent optimizations. We also propose a new optimization-aware decomposition for the inserted SWAP gates. Our experiments show that the routing overhead compiled with our routing algorithm is reduced by up to $69.30\%$ ($21.30\%$ on average) in the number of CNOT gates and up to $43.50\%$ ($7.61\%$ on average) in the circuit depth compared with the state-of-the-art scheme, SABRE.

quant-ph↗