I gave a chat with the workshop on how the synthesis of logic and equipment learning, Specifically places for example statistical relational Finding out, can permit interpretability.
Enthusiastic about synthesizing the semantics of programming languages? We have now a new paper on that, accepted at OOPSLA.
The Lab carries out exploration in artificial intelligence, by unifying Understanding and logic, that has a recent emphasis on explainability
For anyone who is attending NeurIPS this calendar year, you may be interested in checking out our papers that touch on morality, causality, and interpretability. Preprints can be found around the workshop webpage.
We take into account the query of how generalized designs (options with loops) could be considered proper in unbounded and steady domains.
A consortia undertaking on honest systems and goverance was acknowledged late previous 12 months. Information website link below.
We now have a whole new paper approved on Finding out ideal linear programming goals. We get an “implicit“ hypothesis construction approach that yields nice theoretical bounds. Congrats to Gini and Alex https://vaishakbelle.com/ on having this paper recognized. Preprint below.
I gave a seminar on extending the expressiveness of probabilistic relational styles with very first-get features, for instance universal quantification around infinite domains.
Not long ago, he has consulted with major banks on explainable AI and its effect in fiscal establishments.
Along with colleagues from Edinburgh and Herriot Watt, We've got put out the call for a new research agenda.
Paulius' work on algorithmic strategies for randomly building logic packages and probabilistic logic plans continues to be recognized on the ideas and practise of constraint programming (CP2020).
The framework is relevant to a large class of formalisms, like probabilistic relational models. The paper also scientific studies the synthesis challenge in that context. Preprint here.
I gave an invited tutorial the Bathtub CDT Artwork-AI. I protected latest traits and upcoming traits on explainable equipment Discovering.
Conference link Our Focus on symbolically interpreting variational autoencoders, as well as a new learnability for SMT (satisfiability modulo theory) formulation obtained acknowledged at ECAI.