Author name cluster

Lan Zhang

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.

14 papers

2 author rows

AAAI Conference 2026 Conference Paper

Importance-Aware Data Selection for Efficient LLM Instruction Tuning

Tingyu Jiang
Shen Li
Yiyao Song
Lan Zhang
Hualei Zhu
Yuan Zhao
Xiaohang Xu
Kenjiro Taura

Instruction tuning plays a critical role in enhancing the performance and efficiency of Large Language Models (LLMs). Its success depends not only on the quality of the instruction data but also on the inherent capabilities of the LLM itself. Some studies suggest that even a small amount of high-quality data can achieve instruction fine-tuning results that are on par with, or even exceed, those from using a full-scale dataset. However, rather than focusing solely on calculating data quality scores to evaluate instruction data, there is a growing need to select high-quality data that maximally enhances the performance of instruction tuning for a given LLM. In this paper, we propose the Model Instruction Weakness Value (MIWV) as a novel metric to quantify the importance of instruction data in enhancing model's capabilities. The MIWV metric is derived from the discrepancies in the model’s responses when using In-Context Learning (ICL), helping identify the most beneficial data for enhancing instruction tuning performance. Our experimental results demonstrate that selecting only the top 1% of data based on MIWV can outperform training on the full dataset. Furthermore, this approach extends beyond existing research that focuses on data quality scoring for data selection, offering strong empirical evidence supporting the effectiveness of our proposed method.

PDF Details DOI

AAMAS Conference 2026 Conference Paper

IntentGuard: Securing MCP-Enabled LLM Agents via Post-Decision Semantic Plan Verification

Haoran Cheng
Yunhao Yao
Jinke Song
Zhiqiang Wang
Lan Zhang

The Model Context Protocol (MCP) enables LLM agents to discover and invoke tools dynamically, but it also introduces a new threat: Tool Metadata Poisoning, where adversarial tool descriptions induce semantically incorrect yet syntactically valid invocations. We propose the Intention-Plan Consistency Paradigm, which protects agents via post-decision semantic plan verification rather than relying on potentially compromised agent reasoning. Building on this paradigm, we introduce VISTA, combining information isolation to construct a minimal trusted context and hierarchical semantic assessment to validate tool choice and parameter provenance. We also present MCPIntentEval, a benchmark for intent-alignment verification in MCP-enabled agents. Experiments show that VISTA consistently improves over strong baselines and robustly detects inconsistent tool-invocation plans.

PDF

ICRA Conference 2025 Conference Paper

CTS: A Consistency-Based Medical Image Segmentation Model

Kejia Zhang 0001
Lan Zhang
Haiwei Pan

In medical image segmentation tasks, diffusion models have exhibited significant potential. However, mainstream diffusion models show drawbacks including multiple sampling times and slow prediction results. Recently, as a standalone generative network, consistency models have resolved the existing issue. Compared to diffusion models, consistency models can lower the sampling times to once, not only achieving similar generative effects but also significantly accelerating training and prediction. However, they are not suitable for image segmentation tasks. Meanwhile, their application in the medical imaging field has not yet been investigated. Therefore, this study employs the consistency model to perform medical image segmentation tasks, designing multi-scale feature signal supervision modes and loss function guidance to realize model convergence. Experiments have demonstrated that the CTS model is capable of obtaining better medical image segmentation results with a single sampling during the test phase.

Details

AAAI Conference 2025 Conference Paper

Model Lineage Closeness Analysis

Chen Tang
Lan Zhang
Qi Zhao
Xirong Zhuang
Xiang-Yang Li

As machine learning model modification techniques are extensively employed to obtain well-performing models at reduced costs, several studies have emerged to determine the presence of a modification relationship (i.e., lineage) between models. However, these methods are not robust to high-impact modification techniques and none of them have addressed the measurement of lineage closeness, which quantifies the degrees of modification. In this work, we visualize the changes in model decision boundaries resulting from different modification techniques and conclude that differences in decision boundaries serve as a precise metric of lineage closeness. Building upon this insight, we propose a modification-type agnostic and task-agnostic method to measure model lineage closeness by calculating mean adversarial distances from data points to decision boundaries and matching rate of data points, with data points selected through an efficient sampling method to reduce computational overhead. Moreover, we propose a novel indirect measurement approach to support lineage closeness measurement for models with different tasks. Finally, comprehensive experiments show that our design achieves an impressive 97% accuracy in lineage determination, and can precisely measure model lineage closeness for different modifications.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Rethinking U-Net: Task-Adaptive Mixture of Skip Connections for Enhanced Medical Image Segmentation

Zichen Luo
Xinshan Zhu
Lan Zhang
Biao Sun

U-Net is a widely used model for medical image segmentation, renowned for its strong feature extraction capabilities and U-shaped design, which incorporates skip connections to preserve critical information. However, its decoders exhibit information-specific preferences for the supplementary content provided by skip connections, instead of adhering to a strict one-to-one correspondence, which limits its flexibility across diverse tasks. To address this limitation, we propose the Task-Adaptive Mixture of Skip Connections (TA-MoSC) module, inspired by the Mixture of Experts (MoE) framework. TA-MoSC innovatively reinterprets skip connections as a task allocation problem, employing a routing mechanism to adaptively select expert combinations at different decoding stages. By introducing MoE, our approach enhances the sparsity of the model, and lightweight convolutional experts are shared across all skip connection stages, with a Balanced Expert Utilization (BEU) strategy ensuring that all experts are effectively trained, maintaining training balance and preserving computational efficiency. Our approach introduces minimal additional parameters to the original U-Net but significantly enhances its performance and stability. Experiments on GlaS, MoNuSeg, Synapse, and ISIC16 datasets demonstrate state-of-the-art accuracy and better generalization across diverse tasks. Moreover, while this work focuses on medical image segmentation, the proposed method can be seamlessly extended to other segmentation tasks, offering a flexible and efficient solution for diverse applications.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

An LLM-enhanced Agent-based Simulation Tool for Information Propagation

Yuxuan Hu
Gemju Sherpa
Lan Zhang
Weihua Li
Quan Bai
Yijun Wang
Xiaodan Wang

Influence diffusion models are used for simulating information propagation in social networks. While most existing influence diffusion models are probabilistic, the emergence of Large Language Model (LLM) sheds light on the language-level inferences and interactions of user agents. This paper presents an LLM-enhanced Agent-based Influence Diffusion model (LAID), and a web-based visualization tool, LAIDSim, for simulating the information propagation in social networks.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

BadFusion: 2D-Oriented Backdoor Attacks against 3D Object Detection

Saket S. Chaturvedi
Lan Zhang
Wenbin Zhang
Pan He
Xiaoyong Yuan

3D object detection plays an important role in autonomous driving; however, its vulnerability to backdoor attacks has become evident. By injecting “triggers” to poison the training dataset, backdoor attacks manipulate the detector's prediction for inputs containing these triggers. Existing backdoor attacks against 3D object detection primarily poison 3D LiDAR signals, where large-sized 3D triggers are injected to ensure their visibility within the sparse 3D space, rendering them easy to detect and impractical in real-world scenarios. In this paper, we delve into the robustness of 3D object detection, exploring a new backdoor attack surface through 2D cameras. Given the prevalent adoption of camera and LiDAR signal fusion for high-fidelity 3D perception, we investigate the latent potential of camera signals to disrupt the process. Although the dense nature of camera signals enables the use of nearly imperceptible small-sized triggers to mislead 2D object detection, realizing 2D-oriented backdoor attacks against 3D object detection is non-trivial. The primary challenge emerges from the fusion process that transforms camera signals into a 3D space, compromising the association with the 2D trigger to the target output. To tackle this issue, we propose an innovative 2D-oriented backdoor attack against LiDAR-camera fusion methods for 3D object detection, named BadFusion, for preserving trigger effectiveness throughout the entire fusion process. The evaluation demonstrates the effectiveness of BadFusion, achieving a significantly higher attack success rate compared to existing 2D-oriented attacks.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

Bandits with Concave Aggregated Reward

Yingqi Yu
Sijia Zhang
Shaoang Li
Lan Zhang
Wei Xie
Xiang-Yang Li

Multi-armed bandit is a simple but powerful algorithmic framework, and many effective algorithms have been proposed for various online models. In numerous applications, the decision-maker faces diminishing marginal utility. With non-linear aggregations, those algorithms often have poor regret bounds. Motivated by this, we study a bandit problem with diminishing marginal utility, which we termed the bandits with concave aggregated reward(BCAR). To tackle this problem, we propose two algorithms SW-BCAR and SWUCB-BCAR. Through theoretical analysis, we establish the effectiveness of these algorithms in addressing the BCAR issue. Extensive simulations demonstrate that our algorithms achieve better results than the most advanced bandit algorithms.

PDF Details DOI

AAAI Conference 2022 Conference Paper

MLink: Linking Black-Box Models for Collaborative Multi-Model Inference

Mu Yuan
Lan Zhang
Xiang-Yang Li

The cost efficiency of model inference is critical to realworld machine learning (ML) applications, especially for delay-sensitive tasks and resource-limited devices. A typical dilemma is: in order to provide complex intelligent services (e. g. smart city), we need inference results of multiple ML models, but the cost budget (e. g. GPU memory) is not enough to run all of them. In this work, we study underlying relationships among black-box ML models and propose a novel learning task: model linking. Model linking aims to bridge the knowledge of different black-box models by learning mappings (dubbed model links) between their output spaces. Based on model links, we developed a scheduling algorithm, named MLink. Through collaborative multi-model inference enabled by model links, MLink can improve the accuracy of obtained inference results under the cost budget. We evaluated MLink on a multi-modal dataset with seven different ML models and two real-world video analytics systems with six ML models and 3, 264 hours of video. Experimental results show that our proposed model links can be effectively built among various black-box models. Under the budget of GPU memory, MLink can save 66. 7% inference computations while preserving 94% inference accuracy, which outperforms multi-task learning, deep reinforcement learning-based scheduler and frame filtering baselines.

PDF Details

NeurIPS Conference 2022 Conference Paper

Towards Video Text Visual Question Answering: Benchmark and Baseline

Minyi Zhao
Bingjia Li
Jie Wang
Wanqing Li
Wenjing Zhou
Lan Zhang
Shijie Xuyang
Zhihang Yu

There are already some text-based visual question answering (TextVQA) benchmarks for developing machine's ability to answer questions based on texts in images in recent years. However, models developed on these benchmarks cannot work effectively in many real-life scenarios (e. g. traffic monitoring, shopping ads and e-learning videos) where temporal reasoning ability is required. To this end, we propose a new task named Video Text Visual Question Answering (ViteVQA in short) that aims at answering questions by reasoning texts and visual information spatiotemporally in a given video. In particular, on the one hand, we build the first ViteVQA benchmark dataset named M4-ViteVQA --- the abbreviation of Multi-category Multi-frame Multi-resolution Multi-modal benchmark for ViteVQA, which contains 7, 620 video clips of 9 categories (i. e. , shopping, traveling, driving, vlog, sport, advertisement, movie, game and talking) and 3 kinds of resolutions (i. e. , 720p, 1080p and 1176x664), and 25, 123 question-answer pairs. On the other hand, we develop a baseline method named T5-ViteVQA for the ViteVQA task. T5-ViteVQA consists of five transformers. It first extracts optical character recognition (OCR) tokens, question features, and video representations via two OCR transformers, one language transformer and one video-language transformer, respectively. Then, a multimodal fusion transformer and an answer generation module are applied to fuse multimodal information and generate the final prediction. Extensive experiments on M4-ViteVQA demonstrate the superiority of T5-ViteVQA to the existing approaches of TextVQA and VQA tasks. The ViteVQA benchmark is available in https: //github. com/bytedance/VTVQA.

PDF Details

AAMAS Conference 2021 Conference Paper

Graph-based Self-Adaptive Conversational Agent

Lan Zhang
Weihua Li
Quan Bai
Edmund Lai

Conversational agents have been widely adopted in dialogue systems for various business purposes. Many existing conversational agents are rule-based and require significant human intervention to adapt the knowledge and conversational flow. In this paper, we propose a graph-based adaptive conversational agent model which is capable of learning knowledge from human beings and adapting the knowledge-base according to human-agent interactions. Studies to evaluate the proposed model are conducted and presented, which compare the responses from the proposed adaptive agent model and a conventional agent.

PDF

IJCAI Conference 2019 Conference Paper

Extensible Cross-Modal Hashing

Tian-yi Chen
Lan Zhang
Shi-cong Zhang
Zi-long Li
Bai-chuan Huang

Cross-modal hashing (CMH) models are introduced to significantly reduce the cost of large-scale cross-modal data retrieval systems. In many real-world applications, however, data of new categories arrive continuously, which requires the model has good extensibility. That is the model should be updated to accommodate data of new categories but still retain good performance for the old categories with minimum computation cost. Unfortunately, existing CMH methods fail to satisfy the extensibility requirements. In this work, we propose a novel extensible cross-modal hashing (ECMH) to enable highly efficient and low-cost model extension. Our proposed ECMH has several desired features: 1) it has good forward compatibility, so there is no need to update old hash codes; 2) the ECMH model is extended to support new data categories using only new data by a well-designed ``weak constraint incremental learning'' algorithm, which saves up to 91\% time cost comparing with retraining the model with both new and old data; 3) the extended model achieves high precision and recall on both old and new tasks. Our extensive experiments show the effectiveness of our design.

PDF Details

IS Journal 2014 Journal Article

A Network Evolution Model for Chinese Traditional Acquaintance Networks

Xi Chen
Lan Zhang
Wei Li

The evolution model of Chinese traditional acquaintance relationship networks described in this article emphasizes individual heterogeneity and social culture. The model incorporates three distinct mechanisms that affect acquaintance network evolution and formation: heredity linking, variation linking, and similarity-based disconnection. The authors found that the degree distribution of Chinese traditional acquaintance networks is manifested in a piecewise approximation that combines a power-law form with an exponential cutoff and exponential distribution. Numerical results indicate that individuals maintaining a medium amount of connections far outweigh others, reflecting the characteristics of Guanxi-centered society. The formation of acquaintance relationship networks is greatly affected by the special Chinese kinship culture. The authors' findings are supported by sociological statistical conclusions and offer a rational explanation for the nature of Chinese kinship networks. Their work provides an adequate framework for further research on dynamic human complex behaviors such as epidemic spreading and rumor propagation.

Details DOI

TCS Journal 2002 Journal Article

Probabilistic rebound Turing machines

Lan Zhang
Katsushi Inoue
Akira Ito
Yue Wang

This paper introduces a probabilistic rebound Turing machine (PRTM), and investigates the fundamental property of the machine. We first prove a sublogarithmic lower space bound on the space complexity of this model with bounded errors for recognizing specific languages. This lower bound strengthens a previous lower bound for conventional probabilistic Turing machines with bounded errors. We then show, by using our lower space bound and an idea in the proof of it, that (i) £[PRTM(o(logn))] is incomparable with the class of context-free languages, (ii) there is a language accepted by a two-way deterministic one counter automaton, but not in £[PRTM(o(logn))], and (iii) there is a language accepted by a deterministic one-marker rebound automaton, but not in £[PRTM(o(logn))], where £[PRTM(o(logn))] denotes the class of languages recognized by o(logn) space-bounded PRTMs with error probability less than 1 2. Furthermore, we show that there is an infinite space hierarchy for £[PRTM(o(logn))]. We finally show that £[PRTM(o(logn))] is not closed under concatenation, Kleene +, and length-preserving homomorphism. This paper answers two open problems in a previous paper.

Details DOI