I gave a talk at the workshop on how the synthesis of logic and device Finding out, Specifically areas for instance statistical relational Understanding, can help interpretability.
I is going to be offering a tutorial on logic and Discovering with a concentrate on infinite domains at this year's SUM. Link to event here.
I gave a talk entitled "Perspectives on Explainable AI," at an interdisciplinary workshop specializing in setting up belief in AI.
I attended the SML workshop from the Black Forest, and discussed the connections in between explainable AI and statistical relational learning.
Gave a talk this Monday in Edinburgh on the concepts & exercise of device Finding out, covering motivations & insights from our survey paper. Key thoughts lifted provided, the way to: extract intelligible explanations + modify the model to fit changing requires.
I’ll be supplying a talk for the conference on reasonable and liable AI inside the cyber Actual physical techniques session. Owing to Ram & Christian for https://vaishakbelle.com/ your invitation. Backlink to occasion.
We have a fresh paper approved on Discovering optimal linear programming goals. We just take an “implicit“ speculation design approach that yields good theoretical bounds. Congrats to Gini and Alex on receiving this paper recognized. Preprint in this article.
I gave a seminar on extending the expressiveness of probabilistic relational types with 1st-purchase functions, like universal quantification around infinite domains.
Lately, he has consulted with big banks on explainable AI and its effect in economical institutions.
, to permit programs to know a lot quicker and much more correct versions of the world. We are interested in creating computational frameworks that can describe their choices, modular, re-usable
Extended abstracts of our NeurIPS paper (on PAC-Finding out in very first-order logic) and also the journal paper on abstracting probabilistic products was approved to KR's just lately printed study track.
A journal paper on abstracting probabilistic products continues to be approved. The paper research the semantic constraints that enables one to abstract a fancy, low-stage product with a simpler, substantial-amount 1.
The 1st introduces a primary-order language for reasoning about probabilities in dynamical domains, and the next considers the automatic resolving of probability complications laid out in all-natural language.
Our perform (with Giannis) surveying and distilling methods to explainability in machine learning has long been recognized. Preprint below, but the final Edition will probably be on the web and open up accessibility before long.