arXiv subjects

Zhihao Yao

Publications and source records attributed to Zhihao Yao.

15 recordsLinked to original sources

arXiv2026

Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering

Pseudo-labeling based on Optimal Transport (OT) has become an effective mechanism for enhancing short text clustering. Existing OT methods are short in modeling semantic consistencies between samples, which may assign different pseudo-labels to semantically similar samples. These erroneous pseudo-labels can cause the model to produce inferior clusters. This paper proposes a novel short text clustering framework, which remedies the neglect of semantic consistency in existing OT methods, generating reliable pseudo-labels to facilitate clustering. Specifically, the proposed approach first designs an instance-level attention mechanism to capture semantic relationships between samples, which are then integrated into the OT formulation to endow the transport process with neighborhood semantic awareness. By solving the proposed OT formulation, reliable pseudo-labels are obtained that simultaneously account for sample-to-sample semantic consistency and sample-to-cluster global structure information. These pseudo-labels are then used as supervisory signals to guide the model to achieve accurate clustering. Extensive experiments demonstrate that the proposed approach outperforms state-of-the-art methods. The code is available at: \href{https://github.com/YZH0905/CAOT-STC}{https://github.com/YZH0905/CAOT-STC}.

stat.ML↗

arXiv2026

DuoMorph: Synergistic Integration of FDM Printing and Pneumatic Actuation for Shape-Changing Interfaces

We introduce DuoMorph, a design and fabrication method that synergistically integrates Fused Deposition Modeling (FDM) printing and pneumatic actuation to create novel shape-changing interfaces. In DuoMorph, the printed structures and heat-sealed pneumatic elements are mutually designed to actuate and constrain each other, enabling functions that are difficult for either component to achieve in isolation. Moreover, the entire hybrid structure can be fabricated through a single, seamless process using only a standard FDM printer, including both heat-sealing and 3D and 4D printing. In this paper, we define a design space including four primitive categories that capture the fundamental ways in which printed and pneumatic components can interact. To support this process, we present a fabrication method and an accompanying design tool. Finally, we demonstrate the potential of DuoMorph through a series of example applications and performance demonstrations.

cs.HC↗

arXiv2025

Adaptive Backtracking for Privacy Protection in Large Language Models

The preservation of privacy has emerged as a critical topic in the era of artificial intelligence. However, current work focuses on user-oriented privacy, overlooking severe enterprise data leakage risks exacerbated by the Retrieval-Augmented Generation paradigm. To address this gap, our paper introduces a novel objective: enterprise-oriented privacy concerns. Achieving this objective requires overcoming two fundamental challenges: existing methods such as data sanitization severely degrade model performance, and the field lacks public datasets for evaluation. We address these challenges with several solutions. (1) To prevent performance degradation, we propose ABack, a training-free mechanism that leverages a Hidden State Model to pinpoint the origin of a leakage intention and rewrite the output safely. (2) To solve the lack of datasets, we construct PriGenQA, a new benchmark for enterprise privacy scenarios in healthcare and finance. To ensure a rigorous evaluation, we move beyond simple static attacks by developing a powerful adaptive attacker with Group Relative Policy Optimization. Experiments show that against this superior adversary, ABack improves the overall privacy utility score by up to 15\% over strong baselines, avoiding the performance trade-offs of prior methods.

cs.CR↗

arXiv2025

IOCC: Aligning Semantic and Cluster Centers for Few-shot Short Text Clustering

In clustering tasks, it is essential to structure the feature space into clear, well-separated distributions. However, because short text representations have limited expressiveness, conventional methods struggle to identify cluster centers that truly capture each category's underlying semantics, causing the representations to be optimized in suboptimal directions. To address this issue, we propose IOCC, a novel few-shot contrastive learning method that achieves alignment between the cluster centers and the semantic centers. IOCC consists of two key modules: Interaction-enhanced Optimal Transport (IEOT) and Center-aware Contrastive Learning (CACL). Specifically, IEOT incorporates semantic interactions between individual samples into the conventional optimal transport problem, and generate pseudo-labels. Based on these pseudo-labels, we aggregate high-confidence samples to construct pseudo-centers that approximate the semantic centers. Next, CACL optimizes text representations toward their corresponding pseudo-centers. As training progresses, the collaboration between the two modules gradually reduces the gap between cluster centers and semantic centers. Therefore, the model will learn a high-quality distribution, improving clustering performance. Extensive experiments on eight benchmark datasets show that IOCC outperforms previous methods, achieving up to 7.34\% improvement on challenging Biomedical dataset and also excelling in clustering stability and efficiency. The code is available at: https://anonymous.4open.science/r/IOCC-C438.

stat.ME↗

arXiv2025

PyPitfall: Dependency Chaos and Software Supply Chain Vulnerabilities in Python

Python software development heavily relies on third-party packages. Direct and transitive dependencies create a labyrinth of software supply chains. While it is convenient to reuse code, vulnerabilities within these dependency chains can propagate through dependencies, potentially affecting down-stream packages and applications. PyPI, the official Python package repository, hosts many packages and lacks a comprehensive analysis of the prevalence of vulnerable dependencies. This paper introduces PyPitfall, a quantitative analysis of vulnerable dependencies across the PyPI ecosystem. We analyzed the dependency structures of 378,573 PyPI packages and identified 4,655 packages that explicitly require at least one known-vulnerable version and 141,044 packages that permit vulnerable versions within specified ranges. By characterizing the ecosystem-wide dependency landscape and the security impact of transitive dependencies, we aim to raise awareness of Python software supply chain security.

cs.CR↗

arXiv2025

Spiritus: An AI-Assisted Tool for Creating 2D Characters and Animations

This research presents Spiritus, an AI-assisted creation tool designed to streamline 2D character animation creation while enhancing creative flexibility. By integrating natural language processing and diffusion models, users can efficiently transform natural language descriptions into personalized 2D characters and animations. The system employs automated segmentation, layered costume techniques, and dynamic mesh-skeleton binding solutions to support flexible adaptation of complex costumes and additional components. Spiritus further achieves real-time animation generation and efficient animation resource reuse between characters through the integration of BVH data and motion diffusion models. Experimental results demonstrate Spiritus's effectiveness in reducing technical barriers, enhancing creative freedom, and supporting resource universality. Future work will focus on optimizing user experience and further exploring the system's human-computer collaboration potential.

cs.HC↗

arXiv2025

Virtual Co-presenter: Connecting Deaf and Hard-of-hearing Livestreamers and Hearing audience in E-commerce Livestreaming

Deaf and Hard-of-Hearing (DHH) individuals are increasingly participating as livestreamers in China's e-commerce livestreaming industry but face obstacles that limit the scope and diversity of their audience. Our paper examines these challenges and explores a potential solution for connecting the hearing audience to sign language (SL) livestreaming teams with DHH members in e-commerce livestreaming. We interviewed four SL livestreaming team members and 15 hearing audience members to identify information and emotional communication challenges that discourage the hearing audience from continuing to watch SL livestreaming. Based on these findings, we developed a virtual co-presenter demo, which targets SL livestreaming teams with DHH members as users, through a design workshop with six designers, incorporating voice broadcasting with animations. Follow-up evaluations with previous participants provided positive feedback on the virtual co-presenter's potential to address these challenges. We summarize design suggestions on its functionality and interaction design for further refinement to assist SL livestreaming teams with DHH members in reaching a broader hearing audience.

cs.HC↗

arXiv2025

Reliable Pseudo-labeling via Optimal Transport with Attention for Short Text Clustering

Short text clustering has gained significant attention in the data mining community. However, the limited valuable information contained in short texts often leads to low-discriminative representations, increasing the difficulty of clustering. This paper proposes a novel short text clustering framework, called Reliable \textbf{P}seudo-labeling via \textbf{O}ptimal \textbf{T}ransport with \textbf{A}ttention for Short Text Clustering (\textbf{POTA}), that generate reliable pseudo-labels to aid discriminative representation learning for clustering. Specially, \textbf{POTA} first implements an instance-level attention mechanism to capture the semantic relationships among samples, which are then incorporated as a semantic consistency regularization term into an optimal transport problem. By solving this OT problem, we can yield reliable pseudo-labels that simultaneously account for sample-to-sample semantic consistency and sample-to-cluster global structure information. Additionally, the proposed OT can adaptively estimate cluster distributions, making \textbf{POTA} well-suited for varying degrees of imbalanced datasets. Then, we utilize the pseudo-labels to guide contrastive learning to generate discriminative representations and achieve efficient clustering. Extensive experiments demonstrate \textbf{POTA} outperforms state-of-the-art methods. The code is available at: \href{https://github.com/YZH0905/POTA-STC/tree/main}{https://github.com/YZH0905/POTA-STC/tree/main}.

cs.LG↗

arXiv2025

Discriminative Representation learning via Attention-Enhanced Contrastive Learning for Short Text Clustering

Contrastive learning has gained significant attention in short text clustering, yet it has an inherent drawback of mistakenly identifying samples from the same category as negatives and then separating them in the feature space (false negative separation), which hinders the generation of superior representations. To generate more discriminative representations for efficient clustering, we propose a novel short text clustering method, called Discriminative Representation learning via \textbf{A}ttention-\textbf{E}nhanced \textbf{C}ontrastive \textbf{L}earning for Short Text Clustering (\textbf{AECL}). The \textbf{AECL} consists of two modules which are the pseudo-label generation module and the contrastive learning module. Both modules build a sample-level attention mechanism to capture similarity relationships between samples and aggregate cross-sample features to generate consistent representations. Then, the former module uses the more discriminative consistent representation to produce reliable supervision information for assist clustering, while the latter module explores similarity relationships and consistent representations optimize the construction of positive samples to perform similarity-guided contrastive learning, effectively addressing the false negative separation issue. Experimental results demonstrate that the proposed \textbf{AECL} outperforms state-of-the-art methods. If the paper is accepted, we will open-source the code.

cs.LG↗

arXiv2024

Enabling Generative Design Tools with LLM Agents for Mechanical Computation Devices: A Case Study

In the field of Human-Computer Interaction (HCI), interactive devices with embedded mechanical computation are gaining attention. The rise of these cutting-edge devices has created a need for specialized design tools that democratize the prototyping process. While current tools streamline prototyping through parametric design and simulation, they often come with a steep learning curve and may not fully support creative ideation. In this study, we use fluidic computation interfaces as a case study to explore how design tools for such devices can be augmented by Large Language Model agents (LLMs). Integrated with LLMs, the Generative Design Tool (GDT) better understands the capabilities and limitations of new technologies, proposes diverse and practical applications, and suggests designs that are technically and contextually appropriate. Additionally, it generates design parameters for visualizing results and producing fabrication-ready support files. This paper details the GDT's framework, implementation, and performance while addressing its potential and challenges.

cs.HC↗

arXiv2024

Artificial-Intelligence Generated Code Considered Harmful: A Road Map for Secure and High-Quality Code Generation

Generating code via a LLM (rather than writing code from scratch), has exploded in popularity. However, the security implications of LLM-generated code are still unknown. We performed a study that compared the security and quality of human-written code with that of LLM-generated code, for a wide range of programming tasks, including data structures, algorithms, cryptographic routines, and LeetCode questions. To assess code security we used unit testing, fuzzing, and static analysis. For code quality, we focused on complexity and size. We found that LLM can generate incorrect code that fails to implement the required functionality, especially for more complicated tasks; such errors can be subtle. For example, for the cryptographic algorithm SHA1, LLM generated an incorrect implementation that nevertheless compiles. In cases where its functionality was correct, we found that LLM-generated code is less secure, primarily due to the lack of defensive programming constructs, which invites a host of security issues such as buffer overflows or integer overflows. Fuzzing has revealed that LLM-generated code is more prone to hangs and crashes than human-written code. Quality-wise, we found that LLM generates bare-bones code that lacks defensive programming constructs, and is typically more complex (per line of code) compared to human-written code. Next, we constructed a feedback loop that asked the LLM to re-generate the code and eliminate the found issues (e.g., malloc overflow, array index out of bounds, null dereferences). We found that the LLM fails to eliminate such issues consistently: while succeeding in some cases, we found instances where the re-generated, supposedly more secure code, contains new issues; we also found that upon prompting, LLM can introduce issues in files that were issues-free before prompting.

cs.CR↗

arXiv2024

Casper: Prompt Sanitization for Protecting User Privacy in Web-Based Large Language Models

Web-based Large Language Model (LLM) services have been widely adopted and have become an integral part of our Internet experience. Third-party plugins enhance the functionalities of LLM by enabling access to real-world data and services. However, the privacy consequences associated with these services and their third-party plugins are not well understood. Sensitive prompt data are stored, processed, and shared by cloud-based LLM providers and third-party plugins. In this paper, we propose Casper, a prompt sanitization technique that aims to protect user privacy by detecting and removing sensitive information from user inputs before sending them to LLM services. Casper runs entirely on the user's device as a browser extension and does not require any changes to the online LLM services. At the core of Casper is a three-layered sanitization mechanism consisting of a rule-based filter, a Machine Learning (ML)-based named entity recognizer, and a browser-based local LLM topic identifier. We evaluate Casper on a dataset of 4000 synthesized prompts and show that it can effectively filter out Personal Identifiable Information (PII) and privacy-sensitive topics with high accuracy, at 98.5% and 89.9%, respectively.

cs.CR↗

arXiv2024

AI-Gadget Kit: Integrating Swarm User Interfaces with LLM-driven Agents for Rich Tabletop Game Applications

While Swarm User Interfaces (SUIs) have succeeded in enriching tangible interaction experiences, their limitations in autonomous action planning have hindered the potential for personalized and dynamic interaction generation in tabletop games. Based on the AI-Gadget Kit we developed, this paper explores how to integrate LLM-driven agents within tabletop games to enable SUIs to execute complex interaction tasks. After defining the design space of this kit, we elucidate the method for designing agents that can extend the meta-actions of SUIs to complex motion planning. Furthermore, we introduce an add-on prompt method that simplifies the design process for four interaction behaviors and four interaction relationships in tabletop games. Lastly, we present several application scenarios that illustrate the potential of AI-Gadget Kit to construct personalized interaction in SUI tabletop games. We expect to use our work as a case study to inspire research on multi-agent-driven SUI for other scenarios with complex interaction tasks.

cs.HC↗

arXiv2022

Minimizing Trust with Exclusively-Used Physically-Isolated Hardware

Smartphone owners often need to run security-critical programs on the same device as other untrusted and potentially malicious programs. This requires users to trust hardware and system software to correctly sandbox malicious programs, trust that is often misplaced. Our goal is to minimize the number and complexity of hardware and software components that a smartphone owner needs to trust to withstand adversarial inputs. We present a multi-domain hardware design composed of statically-partitioned, physically-isolated trust domains. We introduce a few simple, formally-verified hardware components to enable a program to gain provably exclusive and simultaneous access to both computation and I/O on a temporary basis. To manage this hardware, we present OctopOS, an OS composed of mutually distrustful subsystems. We present a prototype of this machine (hardware and OS) on a CPU-FPGA board and show that it incurs a small hardware cost compared to modern SoCs. For security-critical programs, we show that this machine significantly reduces the required trust compared to mainstream TEEs while achieving decent performance. For normal programs, performance is similar to a legacy machine.

cs.CR↗

arXiv2015

Serifos: Workload Consolidation and Load Balancing for SSD Based Cloud Storage Systems

Achieving high performance in virtualized data centers requires both deploying high throughput storage clusters, i.e. based on Solid State Disks (SSDs), as well as optimally consolidating the workloads across storage nodes. Nowadays, the only practical solution for cloud storage providers to offer guaranteed performance is to grossly over-provision the storage nodes. The current workload scheduling mechanisms used in production do not have the intelligence to optimally allocate block storage volumes based on the performance of SSDs. In this paper, we introduce Serifos, an autonomous performance modeling and load balancing system designed for SSD-based cloud storage. Serifos takes into account the characteristics of the SSD storage units and constructs hardware dependent workload consolidation models. Thus Serifos is able to predict the latency caused by workload interference and the average latency of concurrent workloads. Furthermore, Serifos leverages an I/O load balancing algorithm to dynamically balance the volumes across the cluster. Experimental results indicate that Serifos consolidation model is able to maintain the mean prediction error of around 10% for heterogeneous hardware. As a result of Serifos load balancing, we found that the variance and the maximum average latency are reduced by 82% and 52%, respectively. The supported Service Level Objectives (SLOs) on the testbed improve 43% on average latency, 32% on the maximum read and 63% on the maximum write latency.

cs.DC↗