Author name cluster

Zeyu Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

2 author rows

AAAI Conference 2026 Conference Paper

Constrained Particle Seeking: Solving Diffusion Inverse Problems with Just Forward Passes

Hongkun Dou
Zike Chen
Zeyu Li
Hongjue Li
Lijun Yang
Yue Deng

Diffusion models have gained prominence as powerful generative tools for solving inverse problems due to their ability to model complex data distributions. However, existing methods typically rely on complete knowledge of the forward observation process to compute gradients for guided sampling, limiting their applicability in scenarios where such information is unavailable. In this work, we introduce *Constrained Particle Seeking (CPS)*, a novel gradient-free approach that leverages all candidate particle information to actively search for the optimal particle while incorporating constraints aligned with high-density regions of the unconditional prior. Unlike previous methods that passively select promising candidates, CPS reformulates the inverse problem as a constrained optimization task, enabling more flexible and efficient particle seeking. We demonstrate that CPS can effectively solve both image and scientific inverse problems, achieving results comparable to gradient-based methods while significantly outperforming gradient-free alternatives.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference

Xiang Liu
Zhenheng Tang
Peijie Dong
Zeyu Li
Bo Li
Xuming Hu
Xiaowen Chu

Large Language Models (LLMs) require significant GPU memory when processing long texts, with the key value (KV) cache consuming up to 70\% of total memory during inference. Although existing compression methods reduce memory by evaluating the importance of individual tokens, they overlook critical semantic relationships between tokens, resulting in fragmented context and degraded performance. We introduce \method{}, which fundamentally reimagines KV cache compression by treating semantic chunks - rather than isolated tokens - as basic compression units. This approach preserves complete linguistic structures and contextual integrity, ensuring that essential meaning is retained even under aggressive compression. Our innovation includes a novel layer-wise index reuse technique that exploits the higher cross-layer similarity of preserved indices in \method{}, reducing computational overhead and improving throughput by 26. 5\%. Comprehensive evaluations on challenging benchmarks: LongBench, Needle-In-A-HayStack, GSM8K, and JailbreakV demonstrate that \method{} outperforms state-of-the-art methods by up to 8. 7\% in precision while maintaining the same compression ratio. These results confirm that semantic-aware compression significantly enhances both efficiency and performance for long-context LLM inference, providing a simple yet effective solution to the memory bottleneck problem. \emph{The code is available at \href{https: //github. com/NVIDIA/kvpress}{link}. }

PDF Details

IJCAI Conference 2025 Conference Paper

EVICheck: Evidence-Driven Independent Reasoning and Combined Verification Method for Fact-Checking

Lingxiao Wang
Lei Shi
Feifei Kou
Ligu Zhu
Chen Ma
Pengfei Zhang
Mingying Xu
Zeyu Li

Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) have demonstrated significant potential in automated fact-checking. However, existing methods face limitations in insufficient evidence utilization and lack of explicit verification criteria. Specifically, these approaches aggregate evidence for collective reasoning without independently analyzing each piece, hindering their ability to leverage the available information thoroughly. Additionally, they rely on simple prompts or few-shot learning for verification, which makes truthfulness judgments less reliable, especially for complex claims. To address these limitations, we propose a novel method to enhance evidence utilization and introduce explicit verification criteria, named EVICheck. Our approach independently reasons each evidence piece and synthesizes the results to enable more thorough exploration and enhance interpretability. Additionally, by incorporating fine-grained truthfulness criteria, we make the model's verification process more structured and reliable, especially when handling complex claims. Experimental results on the public RAWFC dataset demonstrate that EVICheck achieves state-of-the-art performance across all evaluation metrics. Our method demonstrates strong potential in fake news verification, significantly improving the accuracy.

PDF Details DOI

ICLR Conference 2025 Conference Paper

Hybrid Regularization Improves Diffusion-based Inverse Problem Solving

Hongkun Dou
Zeyu Li
Jinyang Du
Lijun Yang
Wen Yao 0001
Yue Deng 0001

Diffusion models, recognized for their effectiveness as generative priors, have become essential tools for addressing a wide range of visual challenges. Recently, there has been a surge of interest in leveraging Denoising processes for Regularization (DR) to solve inverse problems. However, existing methods often face issues such as mode collapse, which results in excessive smoothing and diminished diversity. In this study, we perform a comprehensive analysis to pinpoint the root causes of gradient inaccuracies inherent in DR. Drawing on insights from diffusion model distillation, we propose a novel approach called Consistency Regularization (CR), which provides stabilized gradients without the need for ODE simulations. Building on this, we introduce Hybrid Regularization (HR), a unified framework that combines the strengths of both DR and CR, harnessing their synergistic potential. Our approach proves to be effective across a broad spectrum of inverse problems, encompassing both linear and nonlinear scenarios, as well as various measurement noise statistics. Experimental evaluations on benchmark datasets, including FFHQ and ImageNet, demonstrate that our proposed framework not only achieves highly competitive results compared to state-of-the-art methods but also offers significant reductions in wall-clock time and memory consumption.

Details

ICLR Conference 2025 Conference Paper

Physics-aligned field reconstruction with diffusion bridge

Zeyu Li
Hongkun Dou
Shen Fang
Wang Han
Yue Deng 0001
Lijun Yang

The reconstruction of physical fields from sparse measurements is pivotal in both scientific research and engineering applications. Traditional methods are increasingly supplemented by deep learning models due to their efficacy in extracting features from data. However, except for the low accuracy on complex physical systems, these models often fail to comply with essential physical constraints, such as governing equations and boundary conditions. To overcome this limitation, we introduce a novel data-driven field reconstruction framework, termed the Physics-aligned Schr\"{o}dinger Bridge (PalSB). This framework leverages a diffusion bridge mechanism that is specifically tailored to align with physical constraints. The PalSB approach incorporates a dual-stage training process designed to address both local reconstruction mapping and global physical principles. Additionally, a boundary-aware sampling technique is implemented to ensure adherence to physical boundary conditions. We demonstrate the effectiveness of PalSB through its application to three complex nonlinear systems: cylinder flow from Particle Image Velocimetry experiments, two-dimensional turbulence, and a reaction-diffusion system. The results reveal that PalSB not only achieves higher accuracy but also exhibits enhanced compliance with physical constraints compared to existing methods. This highlights PalSB's capability to generate high-quality representations of intricate physical interactions, showcasing its potential for advancing field reconstruction techniques. The source code can be found at https://github.com/lzy12301/PalSB.

Details

IJCAI Conference 2024 Conference Paper

Protecting Object Detection Models from Model Extraction Attack via Feature Space Coverage

Zeyu Li
Yuwen Pu
Xuhong Zhang
Yu Li
Jinbao Li
Shouling Ji

The model extraction attack is an attack pattern aimed at stealing well-trained machine learning models' functionality or privacy information. With the gradual popularization of AI-related technologies in daily life, various well-trained models are being deployed. As a result, these models are considered valuable assets and attractive to model extraction attackers. Currently, the academic community primarily focuses on defense for model extraction attacks in the context of classification, with little attention to the more commonly used task scenario of object detection. Therefore, we propose a detection framework targeting model extraction attacks against object detection models in this paper. The framework first locates suspicious users based on feature coverage in query traffic and uses an active verification module to confirm whether the identified suspicious users are attackers. Through experiments conducted in multiple task scenarios, we validate the effectiveness and detection efficiency of the proposed method.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Should We Really Edit Language Models? On the Evaluation of Edited Language Models

Qi Li
Xiang Liu
Zhenheng Tang
Peijie Dong
Zeyu Li
Xinglin Pan
Xiaowen Chu

Model editing has become an increasingly popular alternative for efficiently updating knowledge within language models. Current methods mainly focus on reliability, generalization, and locality, with many methods excelling across these criteria. Some recent works disclose the pitfalls of these editing methods such as knowledge distortion or conflict. However, the general abilities of post-edited language models remain unexplored. In this paper, we perform a comprehensive evaluation on various editing methods and different language models, and have following findings. (1) Existing editing methods lead to inevitable performance deterioration on general benchmarks, indicating that existing editing methods maintain the general abilities of the model within only a few dozen edits. When the number of edits is slightly large, the intrinsic knowledge structure of the model is disrupted or even completely damaged. (2) Instruction-tuned models are more robust to editing, showing less performance drop on general knowledge after editing. (3) Language model with large scale is more resistant to editing compared to small model. (4) The safety of the edited model, is significantly weakened, even for those safety-aligned models. Our findings indicate that current editing methods are only suitable for small-scale knowledge updates within language models, which motivates further research on more practical and reliable editing methods.

PDF Details DOI

AAAI Conference 2019 Conference Paper

Personalized Question Routing via Heterogeneous Network Embedding

Zeyu Li
Jyun-Yu Jiang
Yizhou Sun
Wei Wang

Question Routing (QR) on Community-based Question Answering (CQA) websites aims at recommending answerers that have high probabilities of providing the “accepted answers” to new questions. The existing question routing algorithms simply predict the ranking of users based on query content. As a consequence, the question raiser information is ignored. On the other hand, they lack learnable scoring functions to explicitly compute ranking scores. To tackle these challenges, we propose NeRank that (1) jointly learns representations of question content, question raiser, and question answerers by a heterogeneous information network embedding algorithm and a long short-term memory (LSTM) model. The embeddings of the three types of entities are unified in the same latent space, and (2) conducts question routing for personalized queries, i. e. , queries with two entities (question content, question raiser), by a convolutional scoring function taking the learned embeddings of all three types of entities as input. Using the scores, NeRank routes new questions to high-ranking answerers that are skillfulness in the question domain and have similar backgrounds to the question raiser. Experimental results show that NeRank significantly outperforms competitive baseline question routing models that ignore the raiser information in three ranking metrics. In addition, NeRank is convergeable in several thousand iterations and insensitive to parameter changes, which prove its effectiveness, scalability, and robustness.

PDF Details