Special Sessions
Special Session Ⅰ: Machine Reasoning
Session Chair: Prof. Ke Qin, University of Electronic Science and Technology of China, China
Session Co-chairs: Assoc. Researcher Shuang Liang, University of Electronic Science and Technology of China, China
Assoc. Researcher Dongyang Zhang, University of Electronic Science and Technology of China, China
Special Session Information:
With the explosive growth of Large Language Models (LLMs) and multimodal technologies, AI is evolving from perception toward cognition. However, the limitations of machine reasoning and the proliferation of generative content pose significant trust and security challenges for artificial intelligence. This session focuses on breakthroughs in LLM reasoning mechanisms, frontiers in multimodal fusion, and detection technologies for generated content. It aims to explore how to build next‑generation AI systems that are more reliable, secure, and capable of stronger reasoning, which is crucial for promoting the robust deployment of AI technologies. This session is intended for researchers, engineers, and graduate students from both academia and industry. Professionals working in natural language processing, computer vision, multimodal learning, AI safety and alignment, and related fields are particularly welcome. Participants will gain insights into the latest advances in machine reasoning and discuss core algorithms and defense strategies for generative content detection. Through cross‑disciplinary exchange, attendees will acquire new ideas for addressing complex reasoning and content security issues in real‑world applications and establish extensive international academic collaborations.
Below is an incomplete list of potential topics to be covered in the Special Session:
•LLM reasoning mechanisms and optimization (e.g., model design and training methods for logical reasoning, causal reasoning, and multi-step reasoning)
•Generative content detection and traceability technologies (e.g., identification algorithms and robustness evaluation for multimodal generative content such as text, images, audio, and video)
•Cutting-edge natural language processing technologies (e.g., knowledge graph, innovative methods and applications of semantic understanding, sentiment analysis, machine translation, and dialogue systems)
•Multimodal information processing and fusion (e.g., cross-modal representation learning, multi-source data alignment, modal conversion, and collaborative reasoning)
•Lightweight technologies for LLMs and multimodal models (e.g., model compression, quantization, pruning, distillation methods, and edge device adaptation)
•AI safety and alignment technologies (e.g., LLM hallucination mitigation, generative content risk prevention and control, and improvement of model interpretability and credibility)
•Practical application cases of related technologies (e.g., landing practices and effect analysis in fields such as intelligent healthcare, intelligent education, and enterprise services)
Special Session Keywords:
•LLM based Optimization
•Generative Content Detection
•Visual Question Answering
•Natural Language Processing
•Multimodal Information Processing
DDL: 2026-07-31
Special Session Ⅱ: Security of Artificial Intelligence and Large Models
Session Chair: Prof. Jie Wang, Shanxi Normal University, China
Special Session Information:
This session aims to bring together top global scholars, industry experts, and policymakers to explore the security challenges and defense strategies arising from the rapid development of AI, particularly Large Language Models (LLMs). The forum will delve into the endogenous security mechanisms of LLMs, risks of data privacy leakage, adversarial attacks and defenses, as well as the detection and governance of AI-generated content. Through keynote speeches, technical sessions, and panel discussions, we seek to build an interdisciplinary platform to foster a trustworthy, reliable, and controllable AI security ecosystem.
Special Session Keywords:
•AI Security
•Large Language Models (LLMs)
•Privacy-Preserving Computation
•Model Robustness
•Adversarial Machine Learning
•Data Poisoning
•Deepfake Detection
•Prompt Injection
Special Session Ⅲ: Artificial Intelligence Security: Foundations, Methods, and Applications
Session Chair: Research Assistant Xia Lei, Communication University of China, China
Session Co-chair: Lecturer Yikun Xu, Communication University of China, China
Special Session Information:
Artificial Intelligence Security has become one of the most active research areas with the rapid advancement of foundation models and large language models. While AI systems have demonstrated remarkable capabilities across a wide range of applications, they are increasingly exposed to security threats such as adversarial attacks, data poisoning, prompt injection, model stealing, privacy leakage, jailbreak attacks, and unreliable model behaviors. Developing secure, trustworthy, explainable, robust, and governable AI systems is therefore essential for the safe and responsible deployment of artificial intelligence. This session aims to bring together researchers and practitioners from academia and industry working on Artificial Intelligence Security. Topics of interest include the theoretical foundations of AI security, adversarial machine learning, large language model security, privacy-preserving machine learning, trustworthy AI, explainable AI, AI security evaluation, AI safety, AI governance, and AI ethics. The session also welcomes research on secure AI applications in cybersecurity, smart manufacturing, autonomous driving, healthcare, smart cities, and other safety-critical domains.
Below is an incomplete list of potential topics to be covered in the Special Session:
•Foundations and Key Technologies of Artificial Intelligence Security
•Security of Large Language Models and Generative AI
•Adversarial Attacks, Defenses, and Robust Machine Learning
•Privacy-Preserving Machine Learning and Federated Learning Security
•Trustworthy AI and Explainable AI
•AI Security Evaluation, Risk Analysis, and AI Red Teaming
•AI Fairness, Ethics, and Governance
•Applications of AI Security in Cybersecurity, Smart Manufacturing, Autonomous Driving, Healthcare, Smart Cities, and Other Safety-Critical Domains
Special Session Keywords:
•Artificial Intelligence Security
•Trustworthy AI
•Large Language Model Security
•Privacy-Preserving Machine Learning
•Adversarial Attacks and Defenses
•AI Fairness and Robustness
•AI Governance
•AI Ethics
•Explainable AI
DDL: 2026-09-30