lacoco-lab

Cognitive Perspectives on Language Models

Course Description: Today’s large language models can show behavior that looks strikingly intelligent. This raises many questions about their relation to, and differences from, human cognition and intelligence. This Blockseminar dives into this rapidly emerging field at the intersection of NLP, language, and cognition.

Instructors: Kate McCurdy. For any questions, please contact me by email: kmccurdy@lst.uni-saarland.de

Time (block seminar): 1-4 pm Monday, Wednesday, Friday; August 10-14, 2026.

In addition, there will be an introductory lecture + coordination session on Tuesday May 12, 14h in the Aquarium, 3rd floor in C7 4.

Room: TBD, likely Aquarium, 3rd floor in C7 4

Registration deadline: Wednesday July 10

Syllabus

Note: The syllabus is subject to change. We will discuss this in the introduction meeting.

Date	Topic	Readings	Presenters
Mon, Aug 10, 2026	Next-word prediction, language, and cognition	Wilcox et al. (2025); McCoy et al. (2024); Ying et al., 2025
Weds, Aug 12, 2026	Cognitive risks of LM use	Chandra et al. (2026); Baumann et al. (2026)
Fri, Aug 14, 2026	Do LMs represent cognitive and/or cultural intelligence?	Farrell et al. (2025); Lupyan and Agüera y Arcas (2026)

Evaluation

There is no final project in this seminar. Grades will be evaluated as follows:

• 60% presentation
• 20% reading summary 1
• 20% reading summary 2

Reading summaries

• Only prepare on days you don’t present
• Format:
  • One bullet point (1-2 sentences) summarizing each assigned paper
  • One paragraph (length up to you, but absolute max 1 page - half a page would be standard) giving an overall synthesis / take-away from the day’s readings; I may randomly call on people to read aloud their synthesis paragraph in class
  • One question for discussion
• Upload to CMS before each seminar session

Presentation guidance

• Clearly identify the main argument and evidence for and against key claims
• Give specific examples of the task (e.g. stimuli from behavioral experiment, inputs and outputs to computational experiment)
• Motivate the argument: why does this matter?
• Contextualize with relevant background material as needed for core concepts
• Give your own critical perspective: how could the study be improved? In what ways do or don’t you agree with arguments made by the authors?