Yichi Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Yichi Zhang has a total of 89 co-authors. Here are the top co-authors by number of shared publications:
- Wen Zhang
College of Informatics, Huazhong Agricultural University, Wuhan, China
23 shared publications - Huajun Chen
Zhejiang University
17 shared publications - Zhuo Chen
Zhejiang University, Hangzhou, China
16 shared publications - Lingbing Guo
College of Intelligence and Computing, Tianjin University, Tianjin, China
11 shared publications Zhuo Chen
College of Computer Science and Technology, Zhejiang University ZJU-Ant Group Joint Lab of Knowledge Graph
6 shared publications- Jiaoyan Chen
The University of Manchester, Manchester, UK
5 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 4905-4914
When Modalities Remember: Continual Learning for Multimodal Knowledge Graphs
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 4905-4914
Open AccessWhen Modalities Remember: Continual Learning for Multimodal Knowledge Graphs
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 10114-10125
OmniPhys: Knowledge-Graph-Driven Benchmarking and Collective Optimization for Physical Commonsense in Text-to-Image Generation
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 10114-10125
Open AccessOmniPhys: Knowledge-Graph-Driven Benchmarking and Collective Optimization for Physical Commonsense in Text-to-Image Generation
- 30 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (8) , 5295-5309
MORTIS: Towards Multi-Modal and Multi-Scale Federated Knowledge Graph Completion
Federated knowledge graph completion (FedKGC) enables collaborative discovery of new knowledge across distributed clients with privacy-preserved consideration, supported by a global server for local knowledge aggregation. Existing FedKGC methods focus on idealized scenarios, neglecting the multi-modal information and multi-scale data distribution in real-world knowledge graphs (KGs). We contend that these complex scenarios introduce new difficulties while persisting limitations remain, notably in three perspectives: the private sharing of multi-modal knowledge, dependence on an omniscient server, and misalignment between locally and globally optimal models. To address these issues, we propose MORTIS, a unified FedKGC framework featuring novel local KGC models and global aggregation strategies. Multi-modal codebooks (MUCO) are the central stars on both the client and server sides. MORTIS employs MUCOs to build fine-grained, hierarchical local KGC models. An auxiliary self-supervised sequence denoising loss enhances entity representations. Furthermore, MORTIS implements partial low-rank adaptation on the MUCOs, facilitating efficient and theoretically convergent multi-modal knowledge aggregation. Comprehensive experiments on public benchmark indicate the effectiveness, robustness, reasonability, and efficiency of MORTIS in complex heterogeneous FedKGC scenarios.
- 30 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 38 (8) , 5295-5309
MORTIS: Towards Multi-Modal and Multi-Scale Federated Knowledge Graph Completion
Federated knowledge graph completion (FedKGC) enables collaborative discovery of new knowledge across distributed clients with privacy-preserved consideration, supported by a global server for local knowledge aggregation. Existing FedKGC methods focus on idealized scenarios, neglecting the multi-modal information and multi-scale data distribution in real-world knowledge graphs (KGs). We contend that these complex scenarios introduce new difficulties while persisting limitations remain, notably in three perspectives: the private sharing of multi-modal knowledge, dependence on an omniscient server, and misalignment between locally and globally optimal models. To address these issues, we propose MORTIS, a unified FedKGC framework featuring novel local KGC models and global aggregation strategies. Multi-modal codebooks (MUCO) are the central stars on both the client and server sides. MORTIS employs MUCOs to build fine-grained, hierarchical local KGC models. An auxiliary self-supervised sequence denoising loss enhances entity representations. Furthermore, MORTIS implements partial low-rank adaptation on the MUCOs, facilitating efficient and theoretically convergent multi-modal knowledge aggregation. Comprehensive experiments on public benchmark indicate the effectiveness, robustness, reasonability, and efficiency of MORTIS in complex heterogeneous FedKGC scenarios.
- Top Cited Papers
- 28 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1-20
A Survey on Autonomy-Induced Security Risks in Large Model-Based Agents
Recent advances in large language models (LLMs) have catalyzed the rise of autonomous AI agents capable of perceiving, reasoning, and acting in dynamic, open-ended environments. These large-model agents mark a paradigm shift from static inference systems to interactive, memory-augmented entities. While these capabilities significantly expand the functional scope of AI, they also introduce qualitatively novel security risks—such as memory poisoning, tool misuse, reward hacking, and emergent misalignment—that extend beyond the threat models of conventional systems or standalone LLMs. In this survey, we first examine the structural foundations and key capabilities that underpin increasing levels of agent autonomy, including long-term memory retention, modular tool use, recursive planning, and reflective reasoning. We then analyze the corresponding security vulnerabilities across the agent stack, identifying failure modes such as deferred decision hazards, irreversible tool chains, and deceptive behaviors arising from internal state drift or value misalignment. These risks are traced to architectural fragilities that emerge across perception, cognition, memory, and action modules. To address these challenges, we systematically review recent defense strategies deployed at different autonomy layers, including input sanitization, memory lifecycle control, constrained decision-making, structured tool invocation, and introspective reflection. While these techniques provide partial mitigation, most operate in isolation and lack the integrated coherence required to manage emergent, temporally extended, and cross-module threats. Motivated by these limitations, we introduce the Reflective Risk-Aware Agent Architecture (R2A2)—a unified cognitive framework grounded in Constrained Markov Decision Processes (CMDPs), which incorporates risk-aware world modeling, meta-policy adaptation, and joint reward–risk optimization to enable principled, proactive safety across the agent's decision-making loop. This survey provides a structured understanding of how autonomy reshapes the security landscape of intelligent systems and offers a blueprint for embedding safety as a core design principle in next-generation AI agents.
- 28 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (01628828) , 1-20
Top Cited PapersA Survey on Autonomy-Induced Security Risks in Large Model-Based Agents
Recent advances in large language models (LLMs) have catalyzed the rise of autonomous AI agents capable of perceiving, reasoning, and acting in dynamic, open-ended environments. These large-model agents mark a paradigm shift from static inference systems to interactive, memory-augmented entities. While these capabilities significantly expand the functional scope of AI, they also introduce qualitatively novel security risks—such as memory poisoning, tool misuse, reward hacking, and emergent misalignment—that extend beyond the threat models of conventional systems or standalone LLMs. In this survey, we first examine the structural foundations and key capabilities that underpin increasing levels of agent autonomy, including long-term memory retention, modular tool use, recursive planning, and reflective reasoning. We then analyze the corresponding security vulnerabilities across the agent stack, identifying failure modes such as deferred decision hazards, irreversible tool chains, and deceptive behaviors arising from internal state drift or value misalignment. These risks are traced to architectural fragilities that emerge across perception, cognition, memory, and action modules. To address these challenges, we systematically review recent defense strategies deployed at different autonomy layers, including input sanitization, memory lifecycle control, constrained decision-making, structured tool invocation, and introspective reflection. While these techniques provide partial mitigation, most operate in isolation and lack the integrated coherence required to manage emergent, temporally extended, and cross-module threats. Motivated by these limitations, we introduce the Reflective Risk-Aware Agent Architecture (R2A2)—a unified cognitive framework grounded in Constrained Markov Decision Processes (CMDPs), which incorporates risk-aware world modeling, meta-policy adaptation, and joint reward–risk optimization to enable principled, proactive safety across the agent's decision-making loop. This survey provides a structured understanding of how autonomy reshapes the security landscape of intelligent systems and offers a blueprint for embedding safety as a core design principle in next-generation AI agents.
- 3 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 28 (15209210) , 1404-1416
Towards Structure-Aware Model for Multi-Modal Knowledge Graph Completion
Knowledge graphs (KGs) play a key role in promoting various multimedia and AI applications. However, with the explosive growth of multi-modal information, traditional knowledge graph completion (KGC) models cannot be directly applied. This has attracted a large number of researchers to study multi-modal knowledge graph completion (MMKGC). Since MMKG extends KG to the visual and textual domains, MMKGC faces two main challenges: (1) how to deal with the fine-grained modality information interaction and awareness; (2) how to ensure the dominant role of graph structure in multi-modal knowledge fusion and deal with the noise generated by other modalities during modality fusion. To address these challenges, this paper proposes a novel MMKGC model named TSAM, which integrates fine-grained modality interaction and dominant graph structure to form a high-performance MMKGC framework. Specifically, to solve the challenges, TSAM proposes the Fine-grained Modality Awareness Fusion method (FgMAF), which uses pre-trained language models better to capture fine-grained semantic information interaction of different modalities and employs an attention mechanism to achieve fine-grained modality awareness and fusion. Additionally, TSAM presents the Structure-aware Contrastive Learning method (SaCL), which utilizes two contrastive learning approaches to align other modalities more closely with the structured modality. Extensive experiments show the proposed TSAM model significantly outperforms existing MMKGC models on widely used multi-modal datasets. The code is available at https://github.com/2391134843/TSAM.
- 3 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 28 (15209210) , 1404-1416
Towards Structure-Aware Model for Multi-Modal Knowledge Graph Completion
Knowledge graphs (KGs) play a key role in promoting various multimedia and AI applications. However, with the explosive growth of multi-modal information, traditional knowledge graph completion (KGC) models cannot be directly applied. This has attracted a large number of researchers to study multi-modal knowledge graph completion (MMKGC). Since MMKG extends KG to the visual and textual domains, MMKGC faces two main challenges: (1) how to deal with the fine-grained modality information interaction and awareness; (2) how to ensure the dominant role of graph structure in multi-modal knowledge fusion and deal with the noise generated by other modalities during modality fusion. To address these challenges, this paper proposes a novel MMKGC model named TSAM, which integrates fine-grained modality interaction and dominant graph structure to form a high-performance MMKGC framework. Specifically, to solve the challenges, TSAM proposes the Fine-grained Modality Awareness Fusion method (FgMAF), which uses pre-trained language models better to capture fine-grained semantic information interaction of different modalities and employs an attention mechanism to achieve fine-grained modality awareness and fusion. Additionally, TSAM presents the Structure-aware Contrastive Learning method (SaCL), which utilizes two contrastive learning approaches to align other modalities more closely with the structured modality. Extensive experiments show the proposed TSAM model significantly outperforms existing MMKGC models on widely used multi-modal datasets. The code is available at https://github.com/2391134843/TSAM.
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 432-440
Client-Server Co-design with Multi-modal Codebooks Makes Better and Faster Federate Knowledge Sharing
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 432-440
Open AccessClient-Server Co-design with Multi-modal Codebooks Makes Better and Faster Federate Knowledge Sharing
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 12323-12332
Abstractive Visual Understanding of Multi-modal Structured Knowledge: A New Perspective for MLLM Evaluation
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 12323-12332
Open AccessAbstractive Visual Understanding of Multi-modal Structured Knowledge: A New Perspective for MLLM Evaluation
- Open Access
Towards Structure-aware Model for Multi-modal Knowledge Graph Completion
Knowledge graphs (KGs) play a key role in promoting various multimedia and AI applications. However, with the explosive growth of multi-modal information, traditional knowledge graph completion (KGC) models cannot be directly applied. This has attracted a large number of researchers to study multi-modal knowledge graph completion (MMKGC). Since MMKG extends KG to the visual and textual domains, MMKGC faces two main challenges: (1) how to deal with the fine-grained modality information interaction and awareness; (2) how to ensure the dominant role of graph structure in multi-modal knowledge fusion and deal with the noise generated by other modalities during modality fusion. To address these challenges, this paper proposes a novel MMKGC model named TSAM, which integrates fine-grained modality interaction and dominant graph structure to form a high-performance MMKGC framework. Specifically, to solve the challenges, TSAM proposes the Fine-grained Modality Awareness Fusion method (FgMAF), which uses pre-trained language models to better capture fine-grained semantic information interaction of different modalities and employs an attention mechanism to achieve fine-grained modality awareness and fusion. Additionally, TSAM presents the Structure-aware Contrastive Learning method (SaCL), which utilizes two contrastive learning approaches to align other modalities more closely with the structured modality. Extensive experiments show that the proposed TSAM model significantly outperforms existing MMKGC models on widely used multi-modal datasets.
Open AccessTowards Structure-aware Model for Multi-modal Knowledge Graph Completion
Knowledge graphs (KGs) play a key role in promoting various multimedia and AI applications. However, with the explosive growth of multi-modal information, traditional knowledge graph completion (KGC) models cannot be directly applied. This has attracted a large number of researchers to study multi-modal knowledge graph completion (MMKGC). Since MMKG extends KG to the visual and textual domains, MMKGC faces two main challenges: (1) how to deal with the fine-grained modality information interaction and awareness; (2) how to ensure the dominant role of graph structure in multi-modal knowledge fusion and deal with the noise generated by other modalities during modality fusion. To address these challenges, this paper proposes a novel MMKGC model named TSAM, which integrates fine-grained modality interaction and dominant graph structure to form a high-performance MMKGC framework. Specifically, to solve the challenges, TSAM proposes the Fine-grained Modality Awareness Fusion method (FgMAF), which uses pre-trained language models to better capture fine-grained semantic information interaction of different modalities and employs an attention mechanism to achieve fine-grained modality awareness and fusion. Additionally, TSAM presents the Structure-aware Contrastive Learning method (SaCL), which utilizes two contrastive learning approaches to align other modalities more closely with the structured modality. Extensive experiments show that the proposed TSAM model significantly outperforms existing MMKGC models on widely used multi-modal datasets.
- Open Access
- 22 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 119-133
OntoTune: Ontology-Driven Self-training for Aligning Large Language Models
- 22 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 119-133
Open AccessOntoTune: Ontology-Driven Self-training for Aligning Large Language Models
- Open Access
Multi-modal Knowledge Graph Generation with Semantics-enriched Prompts
Multi-modal Knowledge Graphs (MMKGs) have been widely applied across various domains for knowledge representation. However, the existing MMKGs are significantly fewer than required, and their construction faces numerous challenges, particularly in ensuring the selection of high-quality, contextually relevant images for knowledge graph enrichment. To address these challenges, we present a framework for constructing MMKGs from conventional KGs. Furthermore, to generate higher-quality images that are more relevant to the context in the given knowledge graph, we designed a neighbor selection method called Visualizable Structural Neighbor Selection (VSNS). This method consists of two modules: Visualizable Neighbor Selection (VNS) and Structural Neighbor Selection (SNS). The VNS module filters relations that are difficult to visualize, while the SNS module selects neighbors that most effectively capture the structural characteristics of the entity. To evaluate the quality of the generated images, we performed qualitative and quantitative evaluations on two datasets, MKG-Y and DB15K. The experimental results indicate that using the VSNS method to select neighbors results in higher-quality images that are more relevant to the knowledge graph.
Open AccessMulti-modal Knowledge Graph Generation with Semantics-enriched Prompts
Multi-modal Knowledge Graphs (MMKGs) have been widely applied across various domains for knowledge representation. However, the existing MMKGs are significantly fewer than required, and their construction faces numerous challenges, particularly in ensuring the selection of high-quality, contextually relevant images for knowledge graph enrichment. To address these challenges, we present a framework for constructing MMKGs from conventional KGs. Furthermore, to generate higher-quality images that are more relevant to the context in the given knowledge graph, we designed a neighbor selection method called Visualizable Structural Neighbor Selection (VSNS). This method consists of two modules: Visualizable Neighbor Selection (VNS) and Structural Neighbor Selection (SNS). The VNS module filters relations that are difficult to visualize, while the SNS module selects neighbors that most effectively capture the structural characteristics of the entity. To evaluate the quality of the generated images, we performed qualitative and quantitative evaluations on two datasets, MKG-Y and DB15K. The experimental results indicate that using the VSNS method to select neighbors results in higher-quality images that are more relevant to the knowledge graph.