While large language models (LLMs) perform well on a variety of language-related tasks, they struggle with tasks that require planning. We apply the existing MuZero algorithm to enhance the planning capabilities of LLMs in dialog settings. MuZero uses a neural network to represent observations into a latent space, and then performs Monte Carlo tree search in the latent space using dynamics learned through self-play. We develop a simulated dialog environment to train the MuZero-based model on conversations with a generative LLM such as DialoGPT. We also investigate modifications to the model architecture, such as replacing the representation network by a transformer pretrained on sentence classification. We evaluate our algorithm on realistic multi-turn dialog planning tasks, such as steering the dialog topic to a predefined goal.
A visit to Stanford University, Microsoft and Google – Professor Dr. Lucie Flek, Professor of Data Science & Language Technologies at b-it, accompanied NRW Minister President Hendrik Wüst on a delegation trip to the USA.
A visit to Stanford University, Microsoft and Google – Professor Dr. Lucie Flek, Professor of Data Science & Language Technologies at b-it, accompanied NRW Minister President Hendrik Wüst on a delegation trip to the USA.
Artificial intelligence (AI) is on the rise. Until now, AI applications generally have “black box” character: How AI arrives at its results remains hidden. Prof. Dr. Jürgen Bajorath, a cheminformatics scientist at b-it, and his team have developed a method that reveals how certain AI applications work in pharmaceutical research. The results are unexpected: the AI programs largely remembered known data and hardly learned specific chemical interactions when predicting drug potency. The results have now been published in “Nature Machine Intelligence”.
This talk explores the theoretical aspects of Conversational Information Seeking (CIS) while combining ongoing interaction log analysis and envisioning future research. This talk begins with the core theories underpinning CIS, providing a foundation for the practical insights that follow. The presentation then explores real-world user engagements through interaction log analysis, revealing key patterns and behaviours. The focus shifts to the horizon of information retrieval, with innovative concepts in immersive information seeking. These visionary ideas represent the future of knowledge access.
Professor Sascha Alda becomes new Scientific Director at the Bonn-Aachen International Center for Information Technology (b-it) on the part of the Hochschule Bonn-Rhein-Sieg and succeeds Professor Nico Hochgeschwender, who is leaving the university.
The Bonn-Aachen International Center for Information Technology (b-it) celebrated its twentieth anniversary these days. Since 2002, it has uniquely combined excellence in research with internationally renowned teaching in three master’s degree programs in computer science. Thus, the b-it qualifies students from all over the world for future-oriented professional fields in the areas of life sciences, media informatics and autonomous systems. With the support of the federal and state governments, an institution with a worldwide reputation has emerged over the past 20 years, which at the same time has an impact on the region with innovative research results and highly qualified graduates.
In the presence of NRW Minister President Hendrik Wüst MdL and NRW Minister of Science Ina Brandes as well as leading actors from business and science, the Lamarr Institute for Machine Learning and Artificial Intelligence was officially opened on September 29, 2022. The leading research institute shapes a new generation of Artificial Intelligence (AI) that is high-performing, sustainable, trustworthy and secure in helping to solve key challenges in industry and society. The Lamarr Institute is one of five university-based AI centers of excellence nationwide that have been receiving permanent funding as of this summer as part of the German government’s AI strategy.
This site uses essential cookies (always active for basic functions) and optional cookies for analytics. We respect your privacy under GDPR – choose to accept or reject non-essential cookies. See our Data Protection for details on your rights.
