Speaker: YE RuVenue: 4-311, Chengjun Court, Zijingang CampusAbstract: When laypeople consult multiple experts—whether in courtrooms, hospitals, or public policy—they often prefer that these experts reach their judgments independently. This intuitive preference for expert autonomy has traditionally been explained by the ‘variety of evidence thesis,’ which claims that agreeing testimonies coming from independent sources are more confirmatory than dependent ones. However, this thesis has been challenged, particularly in contexts involving uncertain source reliability.In this talk, I take a novel approach by shifting focus from the informativeness of agreement among expert testimonies to the informativeness of the entire experiment of consulting experts. I present two main results. First, independence does not in general make the expert-consulting experiment more informative. For some independent group of experts, there exists a correlated counterpart group such that consulting the correlated group is more informative, ceteris paribus. Second, however, when we consider the empirical reality that expert correlation is often ambiguous, and account for decision-makers’ aversion to ambiguity, consulting independent experts will be more informative than consulting correlated ones.

Speaker: SHEN JieVenue: Lecture Hall 206, Haina Complex Building, Zijingang CampusAbstract: Many complex nonlinear systems have intrinsic structures such as energy dissipation or conservation, and/or positivity/maximum principle preserving. It is desirable, sometimes necessary, to preserve these structures in a numerical scheme. I will present some recent advances on using the scalar auxiliary variable (SAV) approach and Lagrange multiplier approach to develop highly efficient and accurate structure preserving schemes for a large class of complex nonlinear systems.These schemes can preserve energy dissipation/conservation as well as other global constraints and/or are positivity/bound preserving, only require solving decoupled linear equations with constant coefficients at each time step and can achieve higher-order accuracy.

Speaker: Mike RobinsonVenue: Research and Study Space, 3rd Floor, Humanities Building, Xixi CampusAbstract: Under the framework of United Nations sustainable development, World Heritage sites must not only achieve their own sustainability but also contribute to broader social well-being. However, in practice, the implementation of these goals has posed numerous challenges and dilemmas for heritage site management. This report will use case studies to illustrate how heritage management requires greater creativity and innovation, even though certain approaches remain controversial.

Speaker: Dr. Tom PegramAbstract: This lecture will examine the role of crisis in global governance, its evolving concept and impact on IR scholarship. It will address the ambiguity of the crisis concept, seen in terms like “permacrisis” and “polycrisis,” and its use in various contexts such as financial instability and climate change. The lecture will introduce a distinction between conjunctural and structural crises, using the 2008 financial crisis and ongoing climate crisis as examples. It will show how different crisis framings affect institutional responses and our understanding of stability and change in global governance. Finally, it will encourage IR scholars to reflect on crisis as an analytical tool and pay more attention to the political and normative aspects of crisis discourse.

Speaker: Edwin LeeansyahVenue: 2A-203, ZJE Building, Haining Campus

Speaker: Xiaochun RongVenue: Lecture Hall 210, Haina Complex Building 2, Zijingang CampusAbstract: Fixing E> 0, a complete Riemannian n-manifold M is called E-collapsed, if every unit ball in M has a volume <E. In Riemannian geometry, interplays between a collapsing geometry and topology has been an important component, and complexities in topology of a collapsed M is linked to a bound on curvature. Around 1980-2000, collapsedmanifolds of bounded sectional curvature was intensively studied by Cheeger-Fukaya-Gromov and many others, which has found several applications. In this talk, we will survey a development in the last decade in investigating collapsed manifolds of Ricci curvatureb ounded below and universal covering of every unit ball in M is not collapsed.

Speaker: TIAN KoumaoVenue: 4-311, Chengjun Court, Zijingang CampusAbstract: There is an argument that the ‘mind’ of another person is not hidden inside the skull, but is ‘seen’ directly in the body’s behavior and interaction with others. This is called ‘direct social perception’ (DSP), a concept proposed in the field of phenomenology. Based on this idea, if AI and robots can achieve behavior indistinguishable from that of humans (beyond a certain level), we are forced to see a ‘subject’ with whom we can interact. In this talk, I will point to the passive generation of an enactive intersubjective loop as the condition that makes DSP possible, and argue that this level diverges from the level of active reasoning when interacting with AI. Then, I believe that artificial life-like ‘fragility’ and ‘precariousness’ are necessary for the interaction with artificial subjects to go one step deeper. Finally, I will consider the future of coexistence with AI, distinguishing between the future in which fragile life-like AI incorporating ‘death’ is born and the future in which it is not.

Speaker: ZHANG JunjieVenue: Room 319/321, ZIBS Building, Haining Campus

Speaker: Minghua ChenVenue: 418, Teaching Building No.4, Yuquan CampusAbstract: The Optimal Power Flow (OPF) problem aims to minimize power generation costs while meeting grid demand and adhering to physical constraints, serving as a pivotal challenge in power system operations, its applications span electricity markets, reliability assessments, and grid modernization for carbon neutrality, with global annual operational costs reaching billions of USD. Traditional iterative solvers struggle with large-scale OPF due to its non-convex and NP hard nature.Recent advances in machine learning (ML)have accelerated OPF solutions by orders of magnitude through techniques like constraint classification, warm-start solutions, and direct load-to-solution mapping, while maintaining accuracy.This tutorial explores cutting-edge ML applications in OPF, including active constraint identification, solver strategy learning, and cross-domain optimization generalization.Future directions for enhancing computational efficiency and robustness under increasing renewable energy uncertainty are also discussed.