arXiv subjects

Yi Liu

Publications and source records attributed to Yi Liu.

At least 289 records · Page 16Linked to original sources

arXiv2025

GBSK: Skeleton Clustering via Granular-ball Computing and Multi-Sampling for Large-Scale Data

To effectively handle clustering task for large-scale datasets, we propose a novel scalable skeleton clustering algorithm, namely GBSK, which leverages the granular-ball technique to capture the underlying structure of data. By multi-sampling the dataset and constructing multi-grained granular-balls, GBSK progressively uncovers a statistical "skeleton" -- a spatial abstraction that approximates the essential structure and distribution of the original data. This strategy enables GBSK to dramatically reduce computational overhead while maintaining high clustering accuracy. In addition, we introduce an adaptive version, AGBSK, with simplified parameter settings to enhance usability and facilitate deployment in real-world scenarios. Extensive experiments conducted on standard computing hardware demonstrate that GBSK achieves high efficiency and strong clustering performance on large-scale datasets, including one with up to 100 million instances across 256 dimensions. Our implementation and experimental results are available at: https://github.com/XFastDataLab/GBSK/.

cs.LG↗

arXiv2025

Rate doubly robust estimation for weighted average treatment effects

The weighted average treatment effect (WATE) defines a versatile class of causal estimands for populations characterized by propensity score weights, including the average treatment effect (ATE), treatment effect on the treated (ATT), on controls (ATC), and for the overlap population (ATO). WATE has broad applicability in social and medical research, as many datasets from these fields align with its framework. However, the literature lacks a systematic investigation into the robustness and efficiency conditions for WATE estimation. Although doubly robust (DR) estimators are well-studied for ATE, their applicability to other WATEs remains uncertain. This paper investigates whether widely used WATEs admit DR or rate doubly robust (RDR) estimators and assesses the role of nuisance function accuracy, particularly with machine learning. Using semiparametric efficient influence function (EIF) theory and double/debiased machine learning (DML), we propose three RDR estimators under specific rate and regularity conditions and evaluate their performance via Monte Carlo simulations. Applications to NHANES data on smoking and blood lead levels, and SIPP data on 401(k) eligibility, demonstrate the methods' practical relevance in medical and social sciences.

stat.ME↗

arXiv2025

Sample Efficient Experience Replay in Non-stationary Environments

Reinforcement learning (RL) in non-stationary environments is challenging, as changing dynamics and rewards quickly make past experiences outdated. Traditional experience replay (ER) methods, especially those using TD-error prioritization, struggle to distinguish between changes caused by the agent's policy and those from the environment, resulting in inefficient learning under dynamic conditions. To address this challenge, we propose the Discrepancy of Environment Dynamics (DoE), a metric that isolates the effects of environment shifts on value functions. Building on this, we introduce Discrepancy of Environment Prioritized Experience Replay (DEER), an adaptive ER framework that prioritizes transitions based on both policy updates and environmental changes. DEER uses a binary classifier to detect environment changes and applies distinct prioritization strategies before and after each shift, enabling more sample-efficient learning. Experiments on four non-stationary benchmarks demonstrate that DEER further improves the performance of off-policy algorithms by 11.54 percent compared to the best-performing state-of-the-art ER methods.

cs.LG↗

arXiv2025

Distributed force element decomposition with method of fundamental solutions

Quantifying the contribution of vortex structures to pressure stress is useful for designing flow control strategies to mitigate low or drag. The traditional force-element method focuses on the contribution of vortex structures to the resultant force. However, the contribution of vortex structures to the distributed force can not be identified. To address this problem, a distributed force element method is proposed. This method projects the Navier-Stokes equation to a divergence-free space and obtain a Poisson equation for pressure. The Green's function is employed to express pressure stress on the solid boundary as a combination of volume and surface integrals. Contributions to the distributed source are divided into acceleration surface elements, vorticity surface elements, convection volume elements. The method of fundamental solutions and singular value decomposition are used to efficiently solve the Green's functions. The distributed force element method is validated using laminar flows around a stationary circular cylinder and oscillating circular cylinder, and three-dimensional laminar and turbulent flows around a sphere. In laminar flow over a circular cylinder, fluctuation of forces mainly originate from volume sources, while in flow over an oscillating circular cylinder, inertial and volume source terms may cancel out, suppressing lift fluctuation. In flows over a sphere, a strong negative volume source in the boundary layer on the sphere creates low-pressure regions, while pressure in the separated shear layer is close to zero.

physics.flu-dyn↗

arXiv2025

Variance Estimation for Weighted Average Treatment Effects

Common variance estimation methods for weighted average treatment effects (WATEs) in observational studies include nonparametric bootstrap and model-based, closed-form sandwich variance estimation. However, the computational cost of bootstrap increases with the size of the data at hand. Besides, some replicates may exhibit random violations of the positivity assumption even when the original data do not. Sandwich variance estimation relies on regularity conditions that may be structurally violated. Moreover, the sandwich variance estimation is model-dependent on the propensity score model, the outcome model, or both; thus it does not have a unified closed-form expression. Recent studies have explored the use of wild bootstrap to estimate the variance of the average treatment effect on the treated (ATT). This technique adopts a one-dimensional, nonparametric, and computationally efficient resampling strategy. In this article, we propose a "post-weighting" bootstrap approach as an alternative to the conventional bootstrap, which helps avoid random positivity violations in replicates and improves computational efficiency. We also generalize the wild bootstrap algorithm from ATT to the broader class of WATEs by providing new justification for correctly accounting for sampling variability from multiple sources under different weighting functions. We evaluate the performance of all four methods through extensive simulation studies and demonstrate their application using data from the National Health and Nutrition Examination Survey (NHANES). Our findings offer several practical recommendations for the variance estimation of WATE estimators.

stat.ME↗

arXiv2025

Avoiding Knowledge Edit Skipping in Multi-hop Question Answering with Guided Decomposition

In a rapidly evolving world where information updates swiftly, knowledge in large language models (LLMs) becomes outdated quickly. Retraining LLMs is not a cost-effective option, making knowledge editing (KE) without modifying parameters particularly necessary. We find that although existing retrieval-augmented generation (RAG)-based KE methods excel at editing simple knowledge, they struggle with KE in multi-hop question answering due to the issue of "edit skipping", which refers to skipping the relevant edited fact in inference. In addition to the diversity of natural language expressions of knowledge, edit skipping also arises from the mismatch between the granularity of LLMs in problem-solving and the facts in the edited memory. To address this issue, we propose a novel Iterative Retrieval-Augmented Knowledge Editing method with guided decomposition (IRAKE) through the guidance from single edited facts and entire edited cases. Experimental results demonstrate that IRAKE mitigates the failure of editing caused by edit skipping and outperforms state-of-the-art methods for KE in multi-hop question answering.

cs.CL↗

arXiv2025

SUDER: Self-Improving Unified Large Multimodal Models for Understanding and Generation with Dual Self-Rewards

Building upon large language models (LLMs), recent large multimodal models (LMMs) unify cross-model understanding and generation into a single framework. However, LMMs still struggle to achieve accurate vision-language alignment, prone to generating text responses contradicting the visual input or failing to follow the text-to-image prompts. Current solutions require external supervision (e.g., human feedback or reward models) and only address unidirectional tasks-either understanding or generation. In this work, based on the observation that understanding and generation are naturally inverse dual tasks, we propose \textbf{SUDER} (\textbf{S}elf-improving \textbf{U}nified LMMs with \textbf{D}ual s\textbf{E}lf-\textbf{R}ewards), a framework reinforcing the understanding and generation capabilities of LMMs with a self-supervised dual reward mechanism. SUDER leverages the inherent duality between understanding and generation tasks to provide self-supervised optimization signals for each other. Specifically, we sample multiple outputs for a given input in one task domain, then reverse the input-output pairs to compute the dual likelihood within the model as self-rewards for optimization. Extensive experimental results on visual understanding and generation benchmarks demonstrate that our method can effectively enhance the performance of the model without any external supervision, especially achieving remarkable improvements in text-to-image tasks.

cs.AI↗

arXiv2025

Onset of vortex shedding in flow past Rankine ovals

The Rankine oval is a classical geometry in potential flow, formed by superimposing a uniform stream with velocity U and a source-sink pair separated by distance 2a with strength m, resulting in a closed stagnation streamline whose shape is governed by the dimensionless parameter Ua/m. Although the Rankine body serves as a cornerstone for the classical theory of potential flow, its behavior in viscous flow remains unexplored. The Rankine oval is streamlined in inviscid flow but behaves as a bluff body in viscous flow. The onset of vortex shedding is a critical phenomenon in flows past a bluff body, mapping the transition from steady to periodic wakes. This study systematically investigates the onset of vortex shedding in Rankine oval flows and its associated fluid dynamics by performing direct numerical simulations of incompressible flow past Rankine ovals over Reynolds numbers from 10 to 200 and Ua/m from 0 to 1. The investigation reveals a linear relationship between Ua/m and the critical Reynolds number. This study further characterizes the lift and drag coefficients and Strouhal number, analyzes the vortex formation, and performs a data-driven dimensional analysis. This analysis identifies the dimensionless quantities and empirical formula that determine St and the friction drag coefficient as a function of Re, independent of Ua/m. For sufficiently large Ua/m, the pressure drag can be estimated using potential flow solutions, enabling reliable predictions of the total drag without numerical simulations. These conclusions collectively provide insights into the fluid dynamics of Rankine ovals across diverse flow conditions.

physics.flu-dyn↗

arXiv2025

Shift Current Anomalous Photovoltaics in a Double Perovskite Ferroelectric

Ferroelectric anomalous photovoltaic (APV) effect, as a fascinating physical conceptual phenomenon, holds significant potentials for new optoelectronic device applications. However, due to the knowledge lacking on the origin and underlying mechanism of ferroelectric APV effect, substantial challenges still remain in exploring new APV-active candidate materials. The emerging shift current model, involving the transfer of photogenerated charges through the displacement of wave functions, has attracted considerable attention for its unique insights into the bulk photovoltaic effect. Here, we present strong APV properties in a high-temperature double perovskite ferroelectric (cyclohexylmethylammonium)$_2$CsAgBiBr$_7$, showing an extremely large above-bandgap photovoltage up to about 40 V. This figure-of-merit is far beyond its bandgap of about 2.3 eV and comparable to the state-of-art molecular ferroelectrics. Strikingly, the shift current model reveals an intrinsic correlation with Cs$^{+}$ cation displacement and provides, for the first time, an explicit explanation for the structural origin of ferroelectric APV activities. Besides, its steady-state APV photocurrent exhibits the unique light-polarization dependence, which endows remarkable polarization-sensitivity with the highest polarization ratios of about 41 among the known 2D single-phase materials. As the unprecedented exploration of ferroelectric APV characteristics illuminated by the shift current mechanism, this finding paves a pathway to assemble new optoelectronic smart devices.

cond-mat.mtrl-sci↗

arXiv2025

LvBench: A Benchmark for Long-form Video Understanding with Versatile Multi-modal Question Answering

Despite remarkable recent progress, existing long-form VideoQA datasets fall short of meeting the criteria for genuine long-form video understanding. This is primarily due to the use of short videos for question curation, and the reliance on limited-length sub-clips as clues to answer those questions. Meanwhile, previous datasets have limited focus on question type and modality. To remedy this, we introduce LvBench, a Long-form video understanding benchmark for versatile multi-modal question-answering. Our LvBench stands out from existing long-form VideoQA datasets through three key characteristics: 1) Extended temporal durations: We consider videos ranging from 70 seconds to 4 hours, covering single-scene, multi-scene, and full-scene contexts. This design accounts for both video and clue lengths, capturing diverse contextual dynamics. 2) Diverse question types and modalities: LvBench introduces six distinct question types that evaluate various perceptual and cognitive capabilities, utilizing both video frames and subtitles. 3) High-quality annotations: We employ rigorous manual labeling by human annotators. Our dataset comprises 20,061 question-answer pairs sourced from 100 carefully selected movies across diverse genres, annotated collaboratively by multiple individuals. Analysis involving various baselines reveals a consistent trend: the performance of all existing methods significantly deteriorates when video and clue length increases. We expect LvBench to serve as a valuable resource for future works on long-form video understanding.

cs.CV↗

arXiv2025

Efficient Detection of Toxic Prompts in Large Language Models

Large language models (LLMs) like ChatGPT and Gemini have significantly advanced natural language processing, enabling various applications such as chatbots and automated content generation. However, these models can be exploited by malicious individuals who craft toxic prompts to elicit harmful or unethical responses. These individuals often employ jailbreaking techniques to bypass safety mechanisms, highlighting the need for robust toxic prompt detection methods. Existing detection techniques, both blackbox and whitebox, face challenges related to the diversity of toxic prompts, scalability, and computational efficiency. In response, we propose ToxicDetector, a lightweight greybox method designed to efficiently detect toxic prompts in LLMs. ToxicDetector leverages LLMs to create toxic concept prompts, uses embedding vectors to form feature vectors, and employs a Multi-Layer Perceptron (MLP) classifier for prompt classification. Our evaluation on various versions of the LLama models, Gemma-2, and multiple datasets demonstrates that ToxicDetector achieves a high accuracy of 96.39\% and a low false positive rate of 2.00\%, outperforming state-of-the-art methods. Additionally, ToxicDetector's processing time of 0.0780 seconds per prompt makes it highly suitable for real-time applications. ToxicDetector achieves high accuracy, efficiency, and scalability, making it a practical method for toxic prompt detection in LLMs.

cs.CR↗

arXiv2025

DyPho-SLAM : Real-time Photorealistic SLAM in Dynamic Environments

Visual SLAM algorithms have been enhanced through the exploration of Gaussian Splatting representations, particularly in generating high-fidelity dense maps. While existing methods perform reliably in static environments, they often encounter camera tracking drift and fuzzy mapping when dealing with the disturbances caused by moving objects. This paper presents DyPho-SLAM, a real-time, resource-efficient visual SLAM system designed to address the challenges of localization and photorealistic mapping in environments with dynamic objects. Specifically, the proposed system integrates prior image information to generate refined masks, effectively minimizing noise from mask misjudgment. Additionally, to enhance constraints for optimization after removing dynamic obstacles, we devise adaptive feature extraction strategies significantly improving the system's resilience. Experiments conducted on publicly dynamic RGB-D datasets demonstrate that the proposed system achieves state-of-the-art performance in camera pose estimation and dense map reconstruction, while operating in real-time in dynamic scenes.

cs.RO↗

arXiv2025

Step-Audio 2 Technical Report

This paper presents Step-Audio 2, an end-to-end multi-modal large language model designed for industry-strength audio understanding and speech conversation. By integrating a latent audio encoder and reasoning-centric reinforcement learning (RL), Step-Audio 2 achieves promising performance in automatic speech recognition (ASR) and audio understanding. To facilitate genuine end-to-end speech conversation, Step-Audio 2 incorporates the generation of discrete audio tokens into language modeling, significantly enhancing its responsiveness to paralinguistic information such as speaking styles and emotions. To effectively leverage the rich textual and acoustic knowledge in real-world data, Step-Audio 2 integrates retrieval-augmented generation (RAG) and is able to call external tools such as web search to mitigate hallucination and audio search to switch timbres. Trained on millions of hours of speech and audio data, Step-Audio 2 delivers intelligence and expressiveness across diverse conversational scenarios. Evaluation results demonstrate that Step-Audio 2 achieves state-of-the-art performance on various audio understanding and conversational benchmarks compared to other open-source and commercial solutions. Please visit https://github.com/stepfun-ai/Step-Audio2 for more information.

cs.CL↗

arXiv2025

TombRaider: Entering the Vault of History to Jailbreak Large Language Models

Warning: This paper contains content that may involve potentially harmful behaviours, discussed strictly for research purposes. Jailbreak attacks can hinder the safety of Large Language Model (LLM) applications, especially chatbots. Studying jailbreak techniques is an important AI red teaming task for improving the safety of these applications. In this paper, we introduce TombRaider, a novel jailbreak technique that exploits the ability to store, retrieve, and use historical knowledge of LLMs. TombRaider employs two agents, the inspector agent to extract relevant historical information and the attacker agent to generate adversarial prompts, enabling effective bypassing of safety filters. We intensively evaluated TombRaider on six popular models. Experimental results showed that TombRaider could outperform state-of-the-art jailbreak techniques, achieving nearly 100% attack success rates (ASRs) on bare models and maintaining over 55.4% ASR against defence mechanisms. Our findings highlight critical vulnerabilities in existing LLM safeguards, underscoring the need for more robust safety defences.

cs.CR↗

arXiv2025

Source Code Summarization in the Era of Large Language Models

To support software developers in understanding and maintaining programs, various automatic (source) code summarization techniques have been proposed to generate a concise natural language summary (i.e., comment) for a given code snippet. Recently, the emergence of large language models (LLMs) has led to a great boost in the performance of code-related tasks. In this paper, we undertake a systematic and comprehensive study on code summarization in the era of LLMs, which covers multiple aspects involved in the workflow of LLM-based code summarization. Specifically, we begin by examining prevalent automated evaluation methods for assessing the quality of summaries generated by LLMs and find that the results of the GPT-4 evaluation method are most closely aligned with human evaluation. Then, we explore the effectiveness of five prompting techniques (zero-shot, few-shot, chain-of-thought, critique, and expert) in adapting LLMs to code summarization tasks. Contrary to expectations, advanced prompting techniques may not outperform simple zero-shot prompting. Next, we investigate the impact of LLMs' model settings (including top\_p and temperature parameters) on the quality of generated summaries. We find the impact of the two parameters on summary quality varies by the base LLM and programming language, but their impacts are similar. Moreover, we canvass LLMs' abilities to summarize code snippets in distinct types of programming languages. The results reveal that LLMs perform suboptimally when summarizing code written in logic programming languages compared to other language types. Finally, we unexpectedly find that CodeLlama-Instruct with 7B parameters can outperform advanced GPT-4 in generating summaries describing code implementation details and asserting code properties. We hope that our findings can provide a comprehensive understanding of code summarization in the era of LLMs.

cs.SE↗

arXiv2025

A Distributed Learned Hash Table

Distributed Hash Tables (DHTs) are pivotal in numerous high-impact key-value applications built on distributed networked systems, offering a decentralized architecture that avoids single points of failure and improves data availability. Despite their widespread utility, DHTs face substantial challenges in handling range queries, which are crucial for applications such as LLM serving, distributed storage, databases, content delivery networks, and blockchains. To address this limitation, we present LEAD, a novel system incorporating learned models within DHT structures to significantly optimize range query performance. LEAD utilizes a recursive machine learning model to map and retrieve data across a distributed system while preserving the inherent order of data. LEAD includes the designs to minimize range query latency and message cost while maintaining high scalability and resilience to network churn. Our comprehensive evaluations, conducted in both testbed implementation and simulations, demonstrate that LEAD achieves tremendous advantages in system efficiency compared to existing range query methods in large-scale distributed systems, reducing query latency and message cost by 80% to 90%+. Furthermore, LEAD exhibits remarkable scalability and robustness against system churn, providing a robust, scalable solution for efficient data retrieval in distributed key-value systems.

cs.NI↗

arXiv2025

A High-Granularity Proton CT Enhanced by Track Discrimination

Proton Computed Tomography (pCT) provides a promising solution to enhance the accuracy of Relative Stopping Power (RSP) required for proton therapy planning. This research introduces a novel high-granularity pCT architecture that incorporates a silicon pixel tracking system and a calorimetric range telescope, which uniquely integrates range telescope functionality with track discrimination capabilities. The Bortfeld function fitting and Convolutional Neural Network (CNN) classifier algorithms are developed and applied for discrimination. In simulation studies, both approaches demonstrate the capability to reduce uncertainty in Water Equivalent Path Length (WEPL) determination for individual proton tracks to below 3~mm. The standard imaging protocol (3.2~mGy, $4\times10^{8}$ protons) achieves sub-millimeter spatial resolution ($\sim$0.5 mm) with sub-1\% RSP accuracy. With proton count requirements reduced by track discrimination, an ultra-low-dose protocol (0.16~mGy, $2\times10^{7}$~protons) is proposed with achieved sub-1\% RSP accuracy and $\sim$1.1~mm spatial resolution in simulation. This low-dose performance significantly expands clinical applicability, particularly for pediatric imaging or frequent imaging scenarios. Furthermore, the target 10 MHz proton detection rate suggests potential for real-time image guidance during radiotherapy. By circumventing the need for ultra-precise energy measurements, this design minimizes hardware complexity and provides a scalable foundation for future pCT systems.

physics.med-ph↗

arXiv2025

Essential norm and Schatten class of Hankel operators between weighted Fock spaces

In this paper, we characterize the essential norm of Hankel operators from a weighted Fock space $F_φ^{p}$ to a weighted Lebesgue space $L_φ^{q}$ for all $1 \leq p, q < \infty$. Additionally, we characterize the Schatten-$h$ class membership for Hankel operators acting from $F_φ^{2}$ to $L_φ^{2}$, where the class is defined by a continuous increasing convex function $h$.

math.FA↗