Yoshua Bengio参会！「大模型安全研讨会2025」开启，4月23日齐聚新加坡

Core Viewpoint - The rapid development of large models in artificial intelligence is reshaping the technological landscape, necessitating a focus on safety, ethics, and responsibility in their application [1][2]. Group 1: Workshop Overview - The Second Large Model Safety Workshop 2025 will take place on April 23, 2025, at JW Marriott South Beach, Singapore, organized by Professor Jun Sun from Singapore Management University [1][30]. - The workshop aims to explore core issues related to large model safety, including technical principles, adversarial attacks, content safety, data privacy, ethical norms, and governance [2]. Group 2: Expert Participation - The workshop will feature distinguished scholars from top global institutions, including Turing Award winner Yoshua Bengio, von Neumann Award winner Christopher D. Manning, and computer security expert Dawn Song [2][12][17][23]. - These experts will discuss the latest research findings on content safety, data security, adversarial defense, risk mitigation strategies, and ethical governance [2]. Group 3: Structure and Goals - The event will consist of nine high-level expert presentations and a deep roundtable discussion, providing a platform for both experienced professionals and newcomers in the field of large models [3][4]. - The workshop aims to balance innovation and safety, ensuring that technological advancements align with ethical standards and social responsibilities [3][4]. Group 4: Broader Impact - This workshop is expected to promote in-depth research on large model safety globally and provide critical references for industry standards and future technological development [4].