Online Location Details: Zoom link will be sent to registrants Open to Public: No Primary Audience: Faculty Event Subject: Diversity/Identity/Inclusive Excellence Event Sponsor: Office of the Provost RSVP Information: https://docs.google.com/forms/d/e/1FAIpQLSfaJ99sGslxTAATnluH1Bj403BADIboky-3Az86pL0esXx9Ig/viewform?usp=dialog Event Contact Email: diversity@tufts.edu The goal of these monthly roundtable discussions is to build a supportive community where faculty can discuss timely issues and strategies. This conversation is being convened by the Office of the Vice Provosts for Institutional Inclusive Excellence, Faculty, Education, and Research, and the Center for the Enhancement of Learning and Teaching.

Event Type: Multifaith Observance Event Sponsor: University Chaplaincy (Jainism) This "Festival of Lamps" celebrates the attainment of Moksa by Lord Mahavira. A burning lamp symbolizes the "light of knowledge," which dispels the darkness of delusion and ignorance.

Building: Alumnae Hall City: Medford, MA 02155 Campus: Medford/Somerville campus Location Details: Alumnae Lounge Open to Public: Yes Primary Audience: Alumni and Friends,Interns and Residents Event Type: Community Engagement Event Subject: Education,Global Engagement,Public Service/Government Event Sponsor Details: Computer Science Speaker Name: Ming Chow RSVP Information: https://forms.gle/NQzkMkeManXXq91Z9 Event Contact Name: Ming Chow Event Contact Email: ming.chow@tufts.edu Event Contact Phone: 617-627-2225 This is a two-hour workshop day on cybersecurity where people can learn from Tufts University faculty, students, community, and industry leaders, in short sessions. This workshop day coincides with National Cybersecurity Awareness Month. Each workshop will provide tangible knowledge that can be used immediately. This community workshop is a partnership between the Department of Computer Science and Tisch College. This event is free and open to the public. Neighbors, community partners, and small business owners in Medford, Somerville, and Boston are welcome to attend. Registration is required.

Building: Goddard Chapel City: Medford, MA 02155 Campus: Medford/Somerville campus Wheelchair Accessible: Yes Open to Public: No Event Subject: Religion/Spirituality Event Sponsor: Tufts University RSVP Information: Tufts University Journaling can be a rich spiritual practice but in our busy days, it can be hard to find time and space to sit down with pen and paper. This drop-in space offers just that—a judgement-free zone where you can drink tea, nibble a snack, and see where the pen takes you. Weekly prompts will be provided but feel free to use this time as you see fit. For students, faculty, and staff, starting September 9.

Learn about Tufts in Paris from the program director and Tufts Global Education staff.

Building: Cabot Intercultural Center City: Medford, MA 02155 Campus: Medford/Somerville campus Location Details: Cabot 702 Open to Public: Yes Link: https://forms.office.com/r/1DA2DnC1rz Today, social scientists increasingly rely on automated tools—like supervised machine learning and, more recently, large language models (LLMs)—to analyze vast amounts of text. For example, these tools can help classify whether a political speech promotes nationalism or detect misinformation in international news. But there is a catch: even when these automated methods are “highly accurate” (e.g., accuracy is more than 90%), the small share of mistakes can distort the final research findings if we treat the machine-generated outputs as error-free. In this talk, Naoki Egami introduces a framework called design-based supervised learning (DSL) that addresses this problem. DSL combines the power of automated annotation with a modest amount of careful human coding by experts. The method uses a doubly robust strategy, ensuring that even when prediction errors are not random, the overall statistical conclusions remain valid. In sum, DSL lets researchers take advantage of cutting-edge tools like LLMs while still producing reliable evidence for policymaking and social science. Egami illustrates the framework with two examples where both the key outcome and the main explanatory variables come from text data. Naoki Egami is an Associate Professor in the Department of Political Science at the Massachusetts Institute of Technology and a faculty affiliate of the Institute for Data, Systems, and Society (IDSS) at MIT. Egami specializes in political methodology and developing statistical methods for questions in political science and the social sciences. Specifically, he works on causal inference and machine learning methods.