This work was accomplished when Liyi Chen was an intern at Alibaba Cloud Computing.

B.1 BaselineGHN:GHN-1 GHNs were designed for NAS, which typically make strong assumptions about the choice of operations and their possible dimensions tomakesearch and learning feasible.

Consider the problem of training deep neural networks on large annotated datasets, such as ImageNet [1]. This problem can be formalized as finding optimal parameters for a given neural networka,parameterized byw,w.r.t.

Neural Information Processing Systems http://nips.cc/

Continual lifelong learning is essential to many applications.

However, in many real-world applications, collecting sufficient exploratory interactions is usually impractical, because online datacollection canbecostlyorevendangerous, suchasinhealthcare [4]andautonomous driving [5]. To address this challenge, offline RL [6, 7] develops a new learning paradigm that trains RL agents only with pre-collected offline datasets and thus can abstract away from the cost of online exploration [8-17].

However, in many real-world applications, collecting sufficient exploratory interactions is usually impractical, because online datacollection canbecostlyorevendangerous, suchasinhealthcare [4]andautonomous driving [5]. To address this challenge, offline RL [6, 7] develops a new learning paradigm that trains RL agents only with pre-collected offline datasets and thus can abstract away from the cost of online exploration [8-17].

Disentanglementhasbeen formalized using a symmetry-centric notion for unstructured spaces, however, graphs have eluded a similarly rigorous treatment. We fill this gap with a new notionofconditional symmetryfordisentanglement, andleveragetoolsfromLie algebras toencode graph properties intosubgroups using suitable adaptations of generative models such as Variational Autoencoders.