Speaker: Prof. Dr. Jasmin Blanchette (Chair of Theoretical Computer Science, LMU München)

Title: Tao's Equational Proof Challenge Accepted

(Joint work with Lydia Kondylidou, LMU München, and Prof. Marijn J.H. Heule, CMU Pittsburgh)

In the context of the Equational Theories Project, Terence Tao posed the challenge of finding alternatives to a complicated 62-step proof found by the Vampire superposition prover. We introduce a proof minimization tool called Krympa. Using a combination of brute force and heuristics, and exploiting both Vampire and the Twee equational prover, the tool reduces the 62-step proof to 20 steps, each corresponding to a rewrite. In an empirical evaluation, it also performs well on 1431 equational problems originating from the same project, reducing in particular a 151-step proof to only 10 steps.

There are two options for attending:

• in presence in room  WE5/05.017 (or WE5/05.003 if too small)
• online in Zoom at: https://uni-bamberg.zoom-x.de/j/66302997167

Meeting ID: 663 0299 7167
Passcode: AISE-SS26

Please send an email to christoph.benzmueller(at)uni-bamberg.de in case you want to attend in presence, so that I can plan for an appropriate room size.

Aufgaben:
- Mitarbeit an Tutorien zu AI-Workflows & Fine-Tuning
- Mitarbeit an Tutorien zu Git, CI/CD, und Containers/Kubernetes
- Betreuung von Übungen

Profil:
- Erfahrung mit ML Modelle (Nutzung/Evaluierung) und Fine-Tuning (z. B. LoRA/PEFT)
- Linux-Grundlagen und Shell-Programmierung
- Git-Basics; Docker/Containers von Vorteil
- Erfahrung mit HPC von Vorteil

Rahmen:
- Beginn: Sept-Okt 2026
- Ende: Feb-März 2026 (je nach Anfangsdatum)
- Umfang: 6 Monate - 33 Std./Monat
- Vergütung nach Uni-Satz (ca. 15-16 Euro/Std je nach Abschluss)

Interesse? Bitte CV, Leistungsübersicht und kurze Beschreibung eurer DevOps bzw. LLM/Fine-Tuning-Erfahrung ASAP an David Fuenmayor (david.fuenmayor(at)uni-bamberg.de).

Am 4. Dezember 2025 findet zudem ein internationaler Workshop zum Themengebiet „Computional Ontology and Metaphysics“ statt. Der Workshop bringt Forscherinnen und Forscher zusammen, die an der Anwendung computergestützter Methoden und Werkzeuge zur Erforschung und Bewertung von Ontologien und metaphysischer Theorien interessiert sind. Das Symposium sowie der Workshop finden in englischer Sprache statt. 

Weitere Informationen, Anmeldung und Programm: https://www.uni-bamberg.de/en/aise/activities/winter-2025/

• Neural-guided superoptimization in ethereum. Inf. Softw. Technol. 186: 107800 (2025) doi: https://doi.org/10.1016/j.infsof.2025.107800
• Secure Optimizations on Ethereum Bytecode Jump-Free Sequences. IEEE Trans. Dependable Secur. Comput. 22(4): 3676-3691 (2025) doi: https://doi.org/10.1109/TDSC.2025.3536803 

Eine weitere Publikation von Lara Lawniczak als Erstautorin, die im Zusammenhang mit ihrer Masterarbeit an der Universität Bamberg und ihrem Studienaufenthalt an der Universität Luxemburg im Jahr 2024 steht, wurde bei der ECAI 2025, der „European Conference on Artificial Intelligence (ECAI)” akzeptiert.

Speaker:
Edward N. Zalta (2025 Johan von Spix Visiting Professor)
Senior Research Scholar
Stanford University
https://mally.stanford.edu/zalta.html

Thursday, May 22, 2025, 16:00 - 18:00 Uhr; Room: U2/00.25

Univis: https://univis.uni-bamberg.de/form?dsc=anew/cal_view&db=current:wiai/bereic/lehrst_3/gastvo&eventdate=2025-05-22&ref=tlecture&sem=2024w

Abstract: When a mathematician asserts "π > 3" or "the set of natural numbers is infinite", what do the symbols "π", "3", and ">" mean, and what are we referring to with the expressions "the set of natural numbers" and "is infinite"?  Of course, the mathematician may simply assume that the expressions "π", "3", "the set of natural numbers" signify mathematical objects, that "is infinite" signifies a property of numbers and sets, and that "<" signifies a relation on numbers. But numbers, sets, properties, and relations are abstract entities and it is not clear how to answer the question "In what sense do these entities exist?", since we can't use (the instruments of) the natural sciences to detect or study such entities.  Instead, metaphysics, which attempts to systematize, and state the laws governing, the abstract entities presupposed by the natural sciences, is the
discipline where this question, and related questions, are studied. But to do metaphysics precisely, we need to deploy the techniques of both logic and computer science. In this talk, I explain how collaboration between researchers at Stanford and Bamberg have developed a metaphysical analysis of the realm of mathematical entities that uses both logical and computational techniques to (a) explain what it is we understand by the symbols and expressions used in mathematics and thereby (b) integrate the 4 key disciplines mentioned in the title.