Friday, September 22, 2023

2023-09-25 / 10:30 ~ 12:00

by 최우진(KAIST)

최근의 생성모델에 관하여 스탠포드대학의 Ermon교수팀에서 NeurIPS2019, ICLR2021에 발표한 아래의 2편의 논문을 집중 리뷰하면서 SDE를 이용한 Generative Modeling의 연구동향과 발전 방향을 심층토의하게 됩니다.

2023-09-26 / 16:30 ~ 17:30

IBS-KAIST 세미나 - 이산수학: Solving Problems in Graph Pebbling using Optimization and Structural Techniques

by Carl R. Yerger(Davidson College)

Graph pebbling is a combinatorial game played on an undirected graph with an initial configuration of pebbles. A pebbling move consists of removing two pebbles from one vertex and placing one pebbling on an adjacent vertex. The pebbling number of a graph is the smallest number of pebbles necessary such that, given any initial configuration of pebbles, at least one pebble can be moved to a specified target vertex. In this talk, we will give a survey of several streams of research in pebbling, including describing a theoretical and computational framework that uses mixed-integer linear programming to obtain bounds for the pebbling numbers of graphs. We will also discuss improvements to this framework through the use of newly proved weight functions that strengthen the weight function technique of Hurlbert. Finally, we will discuss some open extremal problems in pebbling, specifically related to Class 0 graphs and describe how structural graph theoretic techniques such as discharging can be used to obtain results. Collaborators on these projects include Dan Cranson, Dominic Flocco, Luke Postle, Jonad Pulaj, Chenxiao Xue, Marshall Yang, Daniel Zhou.

2023-09-26 / 16:00 ~ 17:00

SAARC 세미나 - SAARC 세미나: Colloquium: Quantum-Classical Correspondence from an Analytic Point of View

by 정인지(서울대학교 수리과학부)

We prove that the twisting in Hamiltonian flows on annular domains, which can be quantified by the differential winding of particles around the center of the annulus, is stable to perturbations. In fact, it is possible to prove the stability of the whole of the lifted dynamics to non-autonomous perturbations, though single particle paths are generically unstable. These all-time stability facts are used to establish a number of results related to the long-time behavior of fluid flows. (Joint work with T. Drivas and T. Elgindi)

2023-09-22 / 14:00 ~ 16:00

IBS-KAIST 세미나 - 수리생물학:

by ()

A key goal of synthetic biology is to establish functional biochemical modules with network-independent properties. Antithetic integral feedback (AIF) is a recently developed control module in which two control species perfectly annihilate each other’s biological activity. The AIF module confers robust perfect adaptation to the steady-state average level of a controlled intracellular component when subjected to sustained perturbations. Recent work has suggested that such robustness comes at the unavoidable price of increased stochastic fluctuations around average levels. We present theoretical results that support and quantify this trade-off for the commonly analyzed AIF variant in the idealized limit with perfect annihilation. However, we also show that this trade-off is a singular limit of the control module: Even minute deviations from perfect adaptation allow systems to achieve effective noise suppression as long as cells can pay the corresponding energetic cost. We further show that a variant of the AIF control module can achieve significant noise suppression even in the idealized limit with perfect adaptation. This atypical configuration may thus be preferable in synthetic biology applications.

Events for the 취소된 행사 포함