BZPEERPapers, people, evidence and research pathsDiscover
Research tools
Network
Opportunities
Account
Source author record
Sergey Ivanov appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.
Catalog footprint
Research graph
Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.
BZPEER is loading the nearby papers, people, topics and institutions for this page.
Published work
This textbook covers principles behind main modern deep reinforcement learning algorithms that achieved breakthrough results in many domains from game AI to robotics. All required theory is explained with proofs using unified notation and emphasize on the differences between different types of algorithms and the reasons why they are constructed the way they are.
Recently there was an increasing interest in applications of graph neural networks in non-Euclidean geometry; however, are non-Euclidean representations always useful for graph learning tasks? For different problems such as node classification and link prediction we compute hyperbolic embeddings and conclude that for tasks that require global prediction consistency it might be useful to use non-Euclidean embeddings, while for other tasks Euclidean models are superior. To do so we first fix an issue of the existing models associated with the optimization process at zero curvature. Current hyperbolic models deal with gradients at the origin in ad-hoc manner, which is inefficient and can lead to numerical instabilities. We solve the instabilities of kappa-Stereographic model at zero curvature cases and evaluate the approach of embedding graphs into the manifold in several graph representation learning tasks.
The neutron's permanent electric dipole moment $d_n$ is constrained to below $3\times10^{-26} e~\text{cm}$ (90% C.L.) [ arXiv:hep-ex/0602020, arXiv:1509.04411 ], by experiments using ultracold neutrons (UCN). We plan to improve this limit by an order of magnitude or more with PanEDM, the first experiment exploiting the ILL's new UCN source SuperSUN. SuperSUN is expected to provide a high density of UCN with energies below 80 neV, implying extended statistical reach with respect to existing sources, for experiments that rely on long storage or spin-precession times. Systematic errors in PanEDM are strongly suppressed by passive magnetic shielding, with magnetic field and gradient drifts at the single fT level. A holding-field homogeneity on the order of $10^{-4}$ is achieved in low residual fields, via a high static damping factor and built-in coil system. No comagnetometer is needed for the first order-of-magnitude improvement in $d_n$, thanks to high magnetic stability and an assortment of sensors outside the UCN storage volumes. PanEDM will be commissioned and upgraded in parallel with SuperSUN, to take full advantage of the source's output in each phase. Commissioning is ongoing in 2019, and a new limit in the mid $10^{-27} e~\text{cm}$ range should be possible with two full reactor cycles of data in the commissioned apparatus.