121 citations · 560 across the 19 of their papers we have counts for
4 papers · 1 filter
Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability
Christopher Frye, Colin Rowat, Ilya Feige
Explaining AI systems is fundamental both to the development of high performing models and to the trust placed in them by their users. The Shapley framework for explainability has…
Binary JUNIPR: an interpretable probabilistic model for discrimination
Anders Andreassen, Ilya Feige, Christopher Frye +1
JUNIPR is an approach to unsupervised learning in particle physics that scaffolds a probabilistic model for jets around their representation as binary trees. Separate JUNIPR models…
Parenting: Safe Reinforcement Learning from Human Input
Christopher Frye, Ilya Feige
Autonomous agents trained via reinforcement learning present numerous safety concerns: reward hacking, negative side effects, and unsafe exploration, among others. In the context o…
Invariant-equivariant representation learning for multi-class data
Ilya Feige
Representations learnt through deep neural networks tend to be highly informative, but opaque in terms of what information they learn to encode. We introduce an approach to probabi…