- From: Filip Ilievski <filip.dbrsk@gmail.com>
- Date: Fri, 22 Dec 2023 17:02:13 +0100
- To: semantic-web@w3.org
- Message-ID: <CANbunvh8=Jquw4jotAD0toPiy-CcWwfTifaFf+XsHYh5juozJw@mail.gmail.com>
Hi all! We have a position open for a Postdoc (8 months, with a possible extension to 11 months; 1.0fte) and/or a paid PhD internship at VU Amsterdam, to study how to measure and advance the ability of AI models to draw systematic analogies between educational materials. Analogical reasoning is a core cognitive skill unique to humans (Penn et al., 2008; Hofstadter, 2001). It can be defined as the ability to perceive and utilize the similarities between situations or events based on (systems of) relations rather than surface similarities (Holyoak, 2012; Gentner et al.). Recent work has shown that analogical reasoning over systems of relations, as opposed to proportional analogies, poses a challenge for state-of-the-art AI, including large language models. The goal of this project will be to use analogies to guide human learning based on knowledge about the skills that a person already has. It will be integrated into the Humane AI project, which entails collaboration with European partners. As part of the process, we intend to develop a benchmark for evaluating analogies in education, together with a method based on identifying structural patterns in the skill descriptions. The work will be tightly coupled with cognitive psychology theories of analogy. We look for candidates with a background in NLP, cognitive psychology, or reasoning. The starting date is as soon as possible - applications will be reviewed on a first-come-first-serve basis. For more information and to express interest, feel free to contact: Filip Ilievski <f.ilievski@vu.nl> Assistant Professor (UD1), Vrije Universiteit (VU) Amsterdam website <http://www.ilievski.info/> | Q100107129 <https://www.wikidata.org/wiki/Q100107129> | scholar <https://scholar.google.com/citations?hl=en&user=4ZScBc0AAAAJ>
Received on Friday, 22 December 2023 16:02:31 UTC