Um in einen lebhaften Austausch mit Interessierten zu treten und die Diskussion rund um das Thema Daten- und Digitalkompetenzen an der Universität Hamburg anzuregen, hat das Projekt DDLitLab die Lecture Series ins Leben gerufen. In den Vorträgen mit anschließender Diskussion beleuchten die Referent:innen Ihre Perspektiven auf die Vermittlung von grundlegenden Datenkompetenzen in Schule und Hochschule. Die Veranstaltung richtet sich sowohl an Mitglieder der UHH als auch externe Gäste.
Zum Vortrag
Referent: Prof. Dr. Chris Biemann, Wissenschaftlicher Direktor HCDS und Professor für Sprachtechnologie am Fachbereich Informatik, UHH
In diesem Vortrag tasten wir uns an die Funktionsweise von Sprachmodellen heran. Wie ging die Entwicklung von einfacher Statistik bis hin zu (künstlich) intelligenten Dialogsystemen? Welche Mechanismen, welche technischen Voraussetzungen und welcher menschlicher Input ist nötig zum Erreichen der momentan verfügbaren Sprachkompetenz auf hohem Niveau? Was machen diese Modelle noch falsch, und was sind die vorhersehbaren Entwicklungen der nächste Jahre?
Melden Sie sich hier an: https://www.isa.uni-hamburg.de/ddlitlab/veranstaltungen/lecture-series-sose-2023.html
Kontakt: Carolin Scharfenberg / Tel: +49 40 42838-8889
Recently, language technology has seen tremendous advancements due to the development and use of language models, large machine learning models pre-trained on large amounts of textual data. However, while how humans express themselves in language and how they perceive language is largely driven by their individual sociodemographic and sociocultural backgrounds, language models still only partially account for these social aspects. In this talk, I argue that we should consider these social factors more when researching and applying language technology. Concretely, I will discuss some of our recent works relating to effectiveness, fairness, and inclusiveness, which will illustrate the critical role of social factors in natural language processing.
The lecture will be followed by a discussion on the topic. The event will be held in English and will take place in person. You can also participate online by registering your participation via mail.
Institutions
This is the first call for participation on the 18th MT Marathon that will take place in Helsinki on August 25-29, 2025. The eighteenth edition of the MT Marathon will be organized by the Language Technology Research group at the University of Helsinki, Finland, with sponsorship of EAMT.
Each Machine Translation Marathon is a week-long gathering of machine translation researchers, developers, students and users featuring:
- MT Lectures and Labs covering the basics and tutorials.
- Keynote Talks from experienced researchers and practitioners.
- Presentations of research and open source tools related to MT.
- Hacking Projects to advance tools or research in one week or start new collaborations.
The registration registration is free of charge for EAMT members
The programme is still under construction.
Projects
We collect and share proposed projects before the Marathon, and the project topics are settled on the first day. Usually, most of the projects actually make it to the final presentation and some continue even (long) after the Marathon. More details will be added later.
Open Session
The MT Marathon will again host an open session with poster presentations related to MT/NLP research and open-source tools. We invite students, developers and researchers to submit short abstracts (1 page) featuring previously published results, open-source tool demos, and work in progress. Abstracts are lightly reviewed for topical scope, and all relevant submissions will be accepted for presentation.
Information about submission procedures will be announced later.
Institutions
Artificial Intelligence is transforming how we approach chemical research and synthesis. By teaching language models to understand and generate the language of chemistry, we have developed complementary AI systems that bridge the gap between computational design and experimental reality.
Our large language model system, ChemCrow, represents one of the first demonstrations of an AI system directly controlling robotic synthesis platforms, successfully executing the synthesis of compounds including organocatalysts and chromophores.
Complementing this, our small language model system, Saturn, currently the most sample-efficient molecular design algorithm, enables precise molecular generation with built-in synthesizability constraints. Saturn’s innovations include direct optimization against retrosynthetic predictions and integration of building block availability, ensuring that generated molecules are practically accessible.
Our work demonstrates how different scales of language models can work together to transform chemical research, from initial molecular design through to physical synthesis, potentially revolutionizing drug discovery, catalysis, and materials development.
Institution
Universität Hamburg
Adeline Scharfenberg
Universität Hamburg
Adeline Scharfenberg
Universität Hamburg
Adeline Scharfenberg