Grounded Neuro-Symbolic AI

Seminar WiSe 2026-2027 – 3 ECTS

WiSe 2026-2027

Wednesday 12:15 - 13:45

A104, Sand 14

Prof. Dr. Martin V. Butz, martin.butzspam prevention@uni-tuebingen.de

1- Format:

• Presentations of papers by students + discussion in group, 1 paper per week.

• Attendance is obligatory. You may have 1 unexcused absence.

• Post questions about the paper before the class – you may miss 2 posts max.

2- When you are presenting:

1. Send your slides with the subject line “Seminar Cognition in Context: slides” latest by SATURDAY

2. 23:59 of the week preceding your presentation to martin.butzspam prevention@uni-tuebingen.de 

3. Receive feedback by Monday; then revise your slides.

4. Present slides in class (aim for 30 minutes).

5. Offer discussion questions.

6. Upload slides and (later) the summary on ILIAS.

3- When you are not presenting:

1. Have a look at the paper before the class! Potentially use an LLM to summarize the paper for you. Ask the LLM several clarification questions.

2. Submit at least 1 interesting observation about the paper content and/or the LLM interaction

3. Participate in class discussion.

4- Some key suggestions for preparing the presentation:

• Use pictures/videos for your presentation.

• Given that you are using LLMs, make sure that slides etc. are comprehensible and use common terminology.

• The key is to give presentations that make the key challenges addressed by the papers accessible as well as the solution approach and the functionality of the involved algorithms and methods.

5- Key suggestions for preparing the summary:

• After class, write a 4-page treatise covering.

• An LLM’s summary of the paper (max one page including your prompt – upload the paper to the LLM)

• Your assessment of the summary’s quality, what has been overlooked, important implications of the paper that you found interesting beyond what the LLM saw…

• A summary of the related perception and discussion that unfolded in class – also in relation to the LLM possibly.

• The summary of the papers should focus on the essence of the contents (main hypothesis / message), your LLM interaction, and the discussion during the seminar.

• Take notes during or directly after the discussion of your paper to be able to summaries the discussion.

6- File names for ILIAS uploads:

• Presentation: PXY_YourName.pdf (where XY is the scheduled week number 01,02,…)

• Summaries: SXY_YourName.pdf

7- Grading:

1. Presentation quality consider comprehensibility, good train of thought throughout the presentation, free speech, contact to the audience, quality of the slides (e.g., slides balanced in contents, graphics, images, proper slide titles).

2. Quality of the summary – short and precise summary of most important message of the paper and the evidence for this message, consistency, accessibility, story line, completeness, integration of aspects considered during the discussion in the seminar.

Some more detailed suggestions to prepare a good paper presentation.