Author name cluster

Jiayi 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.

13 papers

2 author rows

TMLR Journal 2026 Journal Article

A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence

Huan-ang Gao
Jiayi Geng
Wenyue Hua
Mengkang Hu
Xinzhe Juan
Hongzhang Liu
Shilong Liu
Jiahao Qiu

Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks but remain fundamentally static, unable to adapt their internal parameters to novel tasks, evolving knowledge domains, or dynamic interaction contexts. As LLMs are increasingly deployed in open-ended, interactive environments, this static nature has become a critical bottleneck, necessitating agents that can adaptively reason, act, and evolve in real time. This paradigm shift ---from scaling static models to developing self-evolving agents --- has sparked growing interest in architectures and methods enabling continual learning and adaptation from data, interactions, and experiences. This survey provides the first systematic and comprehensive review of self-evolving agents, organizing the field around three foundational dimensions --- what to evolve, when to evolve, and how to evolve. We examine evolutionary mechanisms across agent components (e.g., models, memory, tools, architecture), categorize adaptation methods by stages (e.g., intra-test-time, inter-test-time), and analyze the algorithmic and architectural designs that guide evolutionary adaptation (e.g., scalar rewards, textual feedback, single-agent and multi-agent systems). Additionally, we analyze evaluation metrics and benchmarks tailored for self-evolving agents, highlight applications in domains such as coding, education, and healthcare, and identify critical challenges and research directions in safety, scalability, and co-evolutionary dynamics. By providing a structured framework for understanding and designing self-evolving agents, this survey establishes a roadmap for advancing more adaptive, capable, robust, and versatile agentic systems in both research and real-world deployments, and ultimately sheds light on the realization of Artificial Super Intelligence (ASI) where agents evolve autonomously and perform beyond human-level intelligence across a wide array of tasks.

PDF Details

AAAI Conference 2026 Conference Paper

Multi-modal Dynamic Proxy Learning for Personalized Multiple Clustering

Jinfeng Xu
Zheyu Chen
Shuo Yang
Jinze Li
Ziyue Peng
Zewei Liu
Hewei Wang
Jiayi Zhang

Multiple clustering aims to discover diverse latent structures from different perspectives, yet existing methods generate exhaustive clusterings without discerning user interest, necessitating laborious manual screening. Current multi-modal solutions suffer from static semantic rigidity: predefined candidate words fail to adapt to dataset-specific concepts, and fixed fusion strategies ignore evolving feature interactions. To overcome these limitations, we propose Multi-DProxy, a novel multi-modal dynamic proxy learning framework that leverages cross-modal alignment through learnable textual proxies. Multi-DProxy introduces 1) gated cross-modal fusion that synthesizes discriminative joint representations by adaptively modeling feature interactions. 2) dual-constraint proxy optimization where user interest constraints enforce semantic consistency with domain concepts while concept constraints employ hard example mining to enhance cluster discrimination. 3) dynamic candidate management that refines textual proxies through iterative clustering feedback. Therefore, Multi-DProxy not only effectively captures a user's interest through proxies but also enables the identification of relevant clusterings with greater precision. Extensive experiments demonstrate state-of-the-art performance with significant improvements over existing methods across a broad set of multi-clustering benchmarks.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Scaling Law for Large Wireless Models

Ziheng Liu
Jiayi Zhang
Haoyu Wang
Bokai Xu
Chen Zhang
Yiyang Zhu
Enyu Shi

Emerging from recent advances in foundation models, Large Wireless Models (LWMs) represent a new paradigm of general-purpose intelligence for wireless communications that transcends task-specific engineering. The success of foundation models is critically underpinned by scaling laws, which provide a predictable roadmap for how performance scales with resources. However, established scaling laws from language and vision, charting performance as a power-law of model and dataset sizes, are ill-suited for the wireless domain, as their core formulations cannot model the structured nature of the physical channel. To address this, we propose a novel wireless scaling law that extends the classical formulation by modeling two wireless-native factors: channel heterogeneity and discretization granularity. These two factors reshape scaling behavior via nested linear and power-law relationships, recasting the scaling law's parameters (notably the scaling exponent and irreducible loss) from universal constants into dynamic variables dictated by the physical environment. Our physics-aware formulation reveals two key insights: first, that compute-optimal scaling is not dictated by a fixed model-data ratio but is instead a dynamic function of heterogeneity and granularity, and second, that this dependency is particularly sensitive to granularity, allowing significant performance to be unlocked from existing data simply by refining its resolution. Crucially, this establishes a reliable roadmap for designing powerful yet resource-efficient LWMs, translating theoretical insights into actionable engineering principles. Extensive experiments validate our wireless scaling law, showing a 32.31% prediction accuracy improvement over classical laws in diverse wireless scenarios where they fail.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Atom of Thoughts for Markov LLM Test-Time Scaling

Fengwei Teng
Quan Shi
Zhaoyang Yu
Jiayi Zhang
Yuyu Luo
Chenglin Wu
Zhijiang Guo

Large Language Models (LLMs) achieve superior performance through training-time scaling, and test-time scaling further enhances their capabilities by conducting effective reasoning during inference. However, as the scale of reasoning increases, existing test-time scaling methods suffer from accumulated historical information, which not only wastes computational resources but also interferes with effective reasoning. To address this issue, we observe that complex reasoning can be achieved by solving a series of independent and self-contained subquestions. These subquestions are essentially \textit{atomic questions}, exhibiting the memoryless property similar to Markov processes. Based on this observation, we propose Atom of Thoughts (\our), where each state transition consists of decomposing the current question into a dependency-based directed acyclic graph and contracting its subquestions, forming a simplified question that maintains answer equivalence with the original problem. This answer preservation enables the iterative \textit{decomposition-contraction} process to naturally form a meaningful Markov reasoning process. Furthermore, these atomic states can be seamlessly integrated into existing test-time scaling methods, enabling \our to serve as a plug-in enhancement for improving reasoning capabilities.

PDF Details

JBHI Journal 2025 Journal Article

Attention-Based Q-Space Deep Learning Generalized for Accelerated Diffusion Magnetic Resonance Imaging

Fangrong Zong
Zaimin Zhu
Jiayi Zhang
Xiaofeng Deng
Zhuangzhuang Li
Chuyang Ye
Yong Liu

Diffusion magnetic resonance imaging (dMRI) is a non-invasive method for capturing the microanatomical information of tissues by measuring the diffusion weighted signals along multiple directions, which is widely used in the quantification of microstructures. Obtaining microscopic parameters requires dense sampling in the q space, leading to significant time consumption. The most popular approach to accelerating dMRI acquisition is to undersample the q-space data, along with applying deep learning methods to reconstruct quantitative diffusion parameters. However, the reliance on a predetermined q-space sampling strategy often constrains traditional deep learning-based reconstructions. The present study proposed a novel deep learning model, named attention-based q-space deep learning (aqDL), to implement the reconstruction with variable q-space sampling strategies. The aqDL maps dMRI data from different scanning strategies onto a common feature space by using a series of Transformer encoders. The latent features are employed to reconstruct dMRI parameters via a multilayer perceptron. The performance of the aqDL model was assessed utilizing the Human Connectome Project datasets at varying undersampling numbers. To validate its generalizability, the model was further tested on two additional independent datasets. Our results showed that aqDL consistently achieves the highest reconstruction accuracy at various undersampling numbers, regardless of whether variable or predetermined q-space scanning strategies are employed. These findings suggest that aqDL has the potential to be used on general clinical dMRI datasets.

Details DOI

EAAI Journal 2025 Journal Article

Boosting industrial anomaly detection performance using generated artificial fault data

Kai Wang
Jiayi Zhang
Yishun Liu
Jie Han
Xiaofeng Yuan
Le Zhou

Data-driven anomaly detection aims to learn a decision boundary, enveloping the normal region, and separating normal data from abnormal data. However, industrial data are fairly complex due to varying feedstock and unclear transfer processes and chemical reactions. This means the decision boundary will be very complex and even intractable. In addition, process variables are high-dimensional in modern industrial processes, which strengthens the difficulty of boundary extraction. Generally, the boundary should exactly exceed the outermost samples for precisely drawing normal regions. However, what we have in most situations is just normal data contaminated by unknown noises. Hence, conventional solutions that use statistical analysis to define a normal region result in a not-so-accurate decision boundary where missing alarms occur frequently. In addition to the conventional solution based entirely on historical data, i. e. , passive fault detection (PAD), an alternative detection method, active fault detection (AAD), can circumvent the above problem by stimulating system performance through the intervention of auxiliary signal. While it results in disruption of the normal operation conditions for the process, its method to enhance output performance through additional signals inspires us. In this paper, we resort to the ability of deep neural networks to fit nonlinear data and perform dimension reduction. A fault data generation strategy is proposed and the artificially generated fault data are used to regulate the model training. The new virtual fault data aids in suppressing the decision boundary closest to the outermost periphery. We propose the principles of data generation and form a network structure, implementing information fusion of genuine normal samples and virtual fault samples. Two cases demonstrate the efficiency of the proposed method.

Details DOI

NeurIPS Conference 2025 Conference Paper

Efficient Rectified Flow for Image Fusion

Zirui Wang
Jiayi Zhang
Tianwei Guan
Yuhan Zhou
Xingyuan Li
Minjing Dong
Jinyuan Liu

Image fusion is a fundamental and important task in computer vision, aiming to combine complementary information from different modalities to fuse images. In recent years, diffusion models have made significant developments in the field of image fusion. However, diffusion models often require complex computations and redundant inference time, which reduces the applicability of these methods. To address this issue, we propose RFfusion, an efficient one-step diffusion model for image fusion based on Rectified Flow. We incorporate Rectified Flow into the image fusion task to straighten the sampling path in the diffusion model, achieving one-step sampling without the need for additional training, while still maintaining high-quality fusion results. Furthermore, we propose a task-specific variational autoencoder (VAE) architecture tailored for image fusion, where the fusion operation is embedded within the latent space to further reduce computational complexity. To address the inherent discrepancy between conventional reconstruction-oriented VAE objectives and the requirements of image fusion, we introduce a two-stage training strategy. This approach facilitates the effective learning and integration of complementary information from multi-modal source images, thereby enabling the model to retain fine-grained structural details while significantly enhancing inference efficiency. Extensive experiments demonstrate that our method outperforms other state-of-the-art methods in terms of both inference speed and fusion quality.

PDF Details

IROS Conference 2025 Conference Paper

Multi-Cali Anything: Dense Feature Multi-Frame Structure-from-Motion for Large-Scale Camera Array Calibration

Jinjiang You
Hewei Wang 0001
Yijie Li 0003
Mingxiao Huo
Long Van Tran Ha
Mingyuan Ma
Jinfeng Xu 0003
Jiayi Zhang

Calibrating large-scale camera arrays, such as those in dome-based setups, is time-intensive and typically requires dedicated captures of known patterns. While extrinsics in such arrays are fixed due to the physical setup, intrinsics often vary across sessions due to factors like lens adjustments or temperature changes. In this paper, we propose a dense-feature-driven multi-frame calibration method that refines intrinsics directly from scene data, eliminating the necessity for additional calibration captures. Our approach enhances traditional Structure-from-Motion (SfM) pipelines by introducing an extrinsics regularization term to progressively align estimated extrinsics with ground-truth values, a dense feature reprojection term to reduce keypoint errors by minimizing reprojection loss in the feature space, and an intrinsics variance term for joint optimization across multiple frames. Experiments on the Multiface dataset show that our method achieves nearly the same precision as dedicated calibration processes, and significantly enhances intrinsics and 3D reconstruction accuracy. Fully compatible with existing SfM pipelines, our method provides an efficient and practical plug-and-play solution for large-scale camera setups. Our code is publicly available at: https://github.com/YJJfish/Multi-Cali-Anything

Details

AAAI Conference 2023 Conference Paper

Learning towards Selective Data Augmentation for Dialogue Generation

Xiuying Chen
Mingzhe Li
Jiayi Zhang
Xiaoqiang Xia
Chen Wei
Jianwei Cui
Xin Gao
Xiangliang Zhang

As it is cumbersome and expensive to acquire a huge amount of data for training neural dialog models, data augmentation is proposed to effectively utilize existing training samples. However, current data augmentation techniques on the dialog generation task mostly augment all cases in the training dataset without considering the intrinsic attributes between different cases. We argue that not all cases are beneficial for augmentation task, and the cases suitable for augmentation should obey the following two attributes: (1) low-quality (the dialog model cannot generate a high-quality response for the case), (2) representative (the case should represent the property of the whole dataset). Herein, we explore this idea by proposing a Selective Data Augmentation framework (SDA) for the response generation task. SDA employs a dual adversarial network to select the lowest quality and most representative data points for augmentation in one stage. Extensive experiments conducted on two publicly available datasets, i.e., DailyDialog and OpenSubtitles, show that our framework can improve the response generation performance with respect to various metrics

PDF Details DOI

NeurIPS Conference 2021 Conference Paper

A Bi-Level Framework for Learning to Solve Combinatorial Optimization on Graphs

Runzhong Wang
Zhigang Hua
Gan Liu
Jiayi Zhang
Junchi Yan
Feng Qi
Shuang Yang
Jun Zhou

Combinatorial Optimization (CO) has been a long-standing challenging research topic featured by its NP-hard nature. Traditionally such problems are approximately solved with heuristic algorithms which are usually fast but may sacrifice the solution quality. Currently, machine learning for combinatorial optimization (MLCO) has become a trending research topic, but most existing MLCO methods treat CO as a single-level optimization by directly learning the end-to-end solutions, which are hard to scale up and mostly limited by the capacity of ML models given the high complexity of CO. In this paper, we propose a hybrid approach to combine the best of the two worlds, in which a bi-level framework is developed with an upper-level learning method to optimize the graph (e. g. add, delete or modify edges in a graph), fused with a lower-level heuristic algorithm solving on the optimized graph. Such a bi-level approach simplifies the learning on the original hard CO and can effectively mitigate the demand for model capacity. The experiments and results on several popular CO problems like Directed Acyclic Graph scheduling, Graph Edit Distance and Hamiltonian Cycle Problem show its effectiveness over manually designed heuristics and single-level learning methods.

PDF Details

AIIM Journal 2021 Journal Article

Multiple instance convolutional neural network with modality-based attention and contextual multi-instance learning pooling layer for effective differentiation between borderline and malignant epithelial ovarian tumors

Junming Jian
Wei Xia
Rui Zhang
Xingyu Zhao
Jiayi Zhang
Xiaodong Wu
Yong'ai Li
Jinwei Qiang

Malignant epithelial ovarian tumors (MEOTs) are the most lethal gynecologic malignancies, accounting for 90% of ovarian cancer cases. By contrast, borderline epithelial ovarian tumors (BEOTs) have low malignant potential and are generally associated with a good prognosis. Accurate preoperative differentiation between BEOTs and MEOTs is crucial for determining the appropriate surgical strategies and improving the postoperative quality of life. Multimodal magnetic resonance imaging (MRI) is an essential diagnostic tool. Although state-of-the-art artificial intelligence technologies such as convolutional neural networks can be used for automated diagnoses, their application have been limited owing to their high demand for graphics processing unit memory and hardware resources when dealing with large 3D volumetric data. In this study, we used multimodal MRI with a multiple instance learning (MIL) method to differentiate between BEOT and MEOT. We proposed the use of MAC-Net, a multiple instance convolutional neural network (MICNN) with modality-based attention (MA) and contextual MIL pooling layer (C-MPL). The MA module can learn from the decision-making patterns of clinicians to automatically perceive the importance of different MRI modalities and achieve multimodal MRI feature fusion based on their importance. The C-MPL module uses strong prior knowledge of tumor distribution as an important reference and assesses contextual information between adjacent images, thus achieving a more accurate prediction. The performance of MAC-Net is superior, with an area under the receiver operating characteristic curve of 0. 878, surpassing that of several known MICNN approaches. Therefore, it can be used to assist clinical differentiation between BEOTs and MEOTs.

Details DOI

AAAI Conference 2021 Conference Paper

Reasoning in Dialog: Improving Response Generation by Context Reading Comprehension

Xiuying Chen
Zhi Cui
Jiayi Zhang
Chen Wei
Jianwei Cui
Bin Wang
Dongyan Zhao
Rui Yan

In multi-turn dialog, utterances do not always take the full form of sentences (Carbonell 1983), which naturally makes understanding the dialog context more difficult. However, it is essential to fully grasp the dialog context to generate a reasonable response. Hence, in this paper, we propose to improve the response generation performance by examining the model’s ability to answer a reading comprehension question, where the question is focused on the omitted information in the dialog. Enlightened by the multi-task learning scheme, we propose a joint framework that unifies these two tasks, sharing the same encoder to extract the common and task-invariant features with different decoders to learn taskspecific features. To better fusing information from the question and the dialog history in the encoding part, we propose to augment the Transformer architecture with a memory updater, which is designed to selectively store and update the history dialog information so as to support downstream tasks. For the experiment, we employ human annotators to write and examine a large-scale dialog reading comprehension dataset. Extensive experiments are conducted on this dataset, and the results show that the proposed model brings substantial improvements over several strong baselines on both tasks. In this way, we demonstrate that reasoning can indeed help better response generation and vice versa. We release our large-scale dataset for further research1.

PDF Details

AAAI Conference 2021 Conference Paper

Writing Polishment with Simile: Task, Dataset and A Neural Approach

Jiayi Zhang
Zhi Cui
Xiaoqiang Xia
Yalong Guo
Yanran Li
Chen Wei
Jianwei Cui

A simile is a figure of speech that directly makes a comparison, showing similarities between two different things, e. g. “Reading papers can be dull sometimes, like watching grass grow”. Human writers often interpolate appropriate similes into proper locations of the plain text to vivify their writings. However, none of existing work has explored neural simile interpolation, including both locating and generation. In this paper, we propose a new task of Writing Polishment with Simile (WPS) to investigate whether machines are able to polish texts with similes as we human do. Accordingly, we design a two-staged Locate&Gen model based on transformer architecture. Our model firstly locates where the simile interpolation should happen, and then generates a location-specific simile. We also release a large-scale Chinese Simile (CS) dataset containing 5 million similes with context. The experimental results demonstrate the feasibility of WPS task and shed light on the future research directions towards better automatic text polishment.

PDF Details