Fall Term 2023

The CDS Brownbag HS23 will be dedicated to the topic of Large Language Models for the Behavioral Sciences.

Language models like GPT-4 and co. are set to transform how we conduct research and acquire knowledge. In the HS23 CDS Colloquium, we're diving deep into the exciting opportunities and critical challenges associated with these models. We're keen on understanding how these tools can enhance teaching and research in behavioral sciences, covering a spectrum of applications from clinical and educational realms to many other professional settings. In addition, we'll tackle head-on the crucial issues of bias, inequality, and fairness that come along with these applications. To provide a practical, hands-on approach, we're setting up a series of bootcamps covering the research cycle to better comprehend the practical implications and uses of large language models.  

 

Date  

Time

Topic

Presenter

Location

Tuesday

19-Sep

12:00-13:00

Welcome and Introduction to "LLMs for the Behavioral Sciences”

     Rm 00.010 // Zoom

    Tuesday

    26-Sep

    12:00-13:00

    Journal Club: LLMs for the Behavioral Sciences

    The journal club session's questions can be found on ADAM.

    Literature (available on ADAM):

    Required

    • Hagendorff, T. (2023). Machine psychology: Investigating emergent capabilities and behavior in large language models using psychological methods. arXiv preprint arXiv:2303.13988.
    • Dillion, D., Tandon, N., Gu, Y., & Gray, K. (2023). Can AI language models replace human participants? Trends in Cognitive Sciences27(7), 597–600. doi.org/10.1016/j.tics.2023.04.008

    Optional (recommended)

    • Grossmann, I., Feinberg, M., Parker, D. C., Christakis, N. A., Tetlock, P. E., & Cunningham, W. A. (2023). AI and the transformation of social science research. Science380(6650), 1108–1109. doi.org/10.1126/science.adi1778

    Alexandra BagaïniRm 00.010 // Zoom

    Tuesday

    03-Oct

    12:00-13:00Tutorial: Introduction to using LLMsZak HussainRm 00.010 // Zoom

    Tuesday

    10-Oct

    12:00-13:00Reading & Writing Bootcamp

    Rui Mata

    Loreen Tisdall

    Rm 00.010 // Zoom

    Tuesday

    17-Oct

    12:00-13:00

    Journal Club: Ethics of LLM development and application

    Literature (available on ADAM):

    • Porsdam Mann, S., Earp, B. D., Nyholm, S., Danaher, J., Møller, N., Bowman-Smart, H., Hatherley, J., Koplin, J., Plozza, M., Rodger, D., Treit, P. V., Renard, G., McMillan, J., & Savulescu, J. (2023). Generative AI entails a credit–blame asymmetry. Nature Machine Intelligence5(5), Article 5. https://doi.org/10.1038/s42256-023-00653-1

    Zak Hussain

    Alexandra Bagaïni

    Rm 00.010 // Zoom

    Tuesday

    17-Oct

    18:00-19:30

    Bernoulli Lecture: Peter Dayan

      

    Tuesday

    24-Oct

    12:00-13:00Break Rm 00.010 // Zoom

    Tuesday

    31-Oct

    12:00-13:00Cookies & RYunrui LiuRm 00.010 // Zoom

    Tuesday

    07-Nov

    12:00-13:00

    Coding Bootcamp

    https://www.kaggle.com/learn/python 

    Alexandra Bagaïni

    Zak Hussain

    Rm 00.010 // Zoom

    Friday 

    10-Nov

    Evening"Thinking of AI"

    Rui Mata

    Dirk Wulff

     

    Tuesday

    14-Nov

    12:00-13:00

    Talk: LLMs in Clinical Psychology

    Literature (available on ADAM):

    • Forbes, M. K., Neo, B., Nezami, O. M., Fried, E. I., Faure, K., Michelsen, B., Twose, M., & Dras, M. (2023). Elemental psychopathology: distilling constituent symptoms and patterns of repetition in the diagnostic criteria of the DSM-5. Psychological Medicine, 1–9. doi.org/10.1017/S0033291723002544

    • Franz, P. J., Nook, E. C., Mair, P., & Nock, M. K. (2020). Using Topic Modeling to Detect and Describe Self‐Injurious and Related Content on a Large‐Scale Digital Platform. Suicide and Life-Threatening Behavior50(1), 5–18. doi.org/10.1111/sltb.12569

    Loreen TisdallRm 00.010 // Zoom

    Tuesday

    21-Nov

    12:00-13:00

    Journal Club: LLMs in Clinical Psychology

    Literature (available on ADAM):

    • Coppersmith, G., Dredze, M., Harman, C., & Hollingshead, K. (2015). From ADHD to SAD: Analyzing the Language of Mental Health on Twitter through Self-Reported Diagnoses. Proceedings of the 2nd Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality, 1–10. doi.org/10.3115/v1/W15-1201

    Loreen TisdallRm 00.010 // Zoom

    Tuesday

    28-Nov

    12:00-13:00Presentation Bootcamp

    Alexandra Bagaïni

    Loreen Tisdall

    Rm 00.010 // Zoom

    Tuesday

    05-Dec

    12:00-13:00Student Presentations Rm 00.010 // Zoom

    Tuesday

    12-Dec

    12:00-13:00Student Presentations Rm 00.010 // Zoom

    Tuesday

    19-Dec

    12:00-13:00Student Presentations Rm 00.010 // Zoom

    Brown BagThe CDS brownbag (colloquium) takes place in the form of weekly one-hour group meetings and is a forum for discussion and training for MSc students and CDS staff members. 

    Hackathons: MSc students and CDS members are welcome to participate in programming hackathons aimed at developing data-analysis skills (data visualization, programming, statistical modeling). Whenever necessary and desired, students can receive prior training (e.g., during the CDS brown-bag colloquium or through additional workshops) to ensure a productive hackathon. Hackathons can cover different areas of current interest and past examples include hackathons aimed at exploring the topics of diversity, sustainability, or risk perception

    If you have any questions related to CDS events, please contact Laura Wiles.

    To see previous CDS events, please go to previous events.


    Other events

    MSc students and CDS members are also invited to take part in sessions and workshops of related interdisciplinary networks. Please check their individual websites for details.