I gave a talk on the workshop on how the synthesis of logic and equipment learning, In particular places such as statistical relational Mastering, can enable interpretability.
Keen on synthesizing the semantics of programming languages? We have a fresh paper on that, acknowledged at OOPSLA.
The Lab carries out investigate in artificial intelligence, by unifying Studying and logic, with a new emphasis on explainability
Should you be attending NeurIPS this year, you may have an interest in checking out our papers that contact on morality, causality, and interpretability. Preprints can be found over the workshop website page.
Gave a chat this Monday in Edinburgh to the concepts & observe of equipment Discovering, masking motivations & insights from our study paper. Critical issues elevated involved, ways to: extract intelligible explanations + modify the design to fit altering requirements.
A consortia venture on honest techniques and goverance was accepted late very last year. News hyperlink below.
The get the job done is inspired by the necessity to examination and evaluate inference algorithms. A combinatorial argument for that correctness from the Thoughts can be deemed. Preprint in this article.
I gave a seminar on extending the expressiveness of probabilistic relational models with initial-purchase functions, for example common quantification around infinite domains.
Website link In the final 7 days of Oct, I gave a talk informally speaking about explainability and moral accountability in synthetic intelligence. Thanks to the organizers for your invitation.
Jonathan’s paper considers a lifted approached to weighted model integration, like circuit design. Paulius’ paper https://vaishakbelle.com/ develops a measure-theoretic viewpoint on weighted design counting and proposes a means to encode conditional weights on literals analogously to conditional probabilities, which leads to considerable performance improvements.
For the University of Edinburgh, he directs a study lab on synthetic intelligence, specialising inside the unification of logic and equipment Mastering, by using a recent emphasis on explainability and ethics.
The framework is applicable to a large course of formalisms, including probabilistic relational styles. The paper also scientific studies the synthesis problem in that context. Preprint below.
Should you be attending AAAI this 12 months, you might be interested in looking at our papers that contact on fairness, abstraction and generalized sum-product or service issues.
Our paper on synthesizing ideas with loops inside the presence of probabilistic sound, acknowledged the journal of approximate reasoning, has also been acknowledged for the ICAPS journal monitor. Preprint to the total paper right here.