BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:CECAM workshop: "Generative models for classical and quantum matte
 r"
DTSTART;VALUE=DATE:20241202
DTSTAMP:20260527T190528Z
UID:55f2b8e6a7125848b1e9e6514127e85e44886d6d1b9fd82d28fa09c1
CATEGORIES:Conferences - Seminars
DESCRIPTION:You can apply to participate and find all the relevant informa
 tion (speakers\, abstracts\, program\,...) on the event website: https://
 www.cecam.org/workshop-details/generative-models-for-classical-and-quantum
 -matter-1274.\n\nRegistration is required to attend the full event\, take 
 part in the social activities and present a poster at the poster session (
 if any).  However\, the EPFL community is welcome to attend specific
  lectures without registration if the topic is of interest to their resea
 rch. Do not hesitate to contact the CECAM Event Manager if you have any 
 question.\n\nDescription\nGenerative modeling\, a broad term that encompas
 ses many machine learning techniques to generate random variables that res
 emble samples from a given target distribution\, are being heavily explore
 d in the computational molecular sciences. For example\, pioneering works 
 using these techniques in a chemical context include efforts to sample the
 rmal distributions of molecular systems [1]\, enhance free energy methods 
 [2]\, evaluate intractable integrals arising in quantum field theories [3]
 \, carrying out direct simulations\, among many other applications. The ma
 in paradigms that have been exploited in the molecular sciences include li
 kelihood estimation with normalizing flows (i.e.\, training from knowledge
  of the energy alone) or training from an existing dataset\, typically col
 lected using molecular dynamics simulations [4]. At present\, it is not cl
 ear whether substantive accelerations relative to explicit simulations can
  be obtained for every system [5]. \nIndeed\, while there is palpable exc
 itement about generative modeling techniques\, a number of determinative q
 uestions remain:\n \n\n	How do we build generative models for chemical sy
 stems that generalize beyond a training dataset? \n	Which generative mode
 ls are most appropriate for a given task? \n	How do we train transferable
  models that are appropriate for a large class of systems?\n	Which criteri
 a or metrics should we use to evaluate the quality of generative models?\n
 \n \nThis workshop seeks to highlight and explore these questions by brin
 ging together a group of computational scientists with diverse expertise. 
 In particular\, we hope to integrate discussion of neural network architec
 tures that are well-suited to chemical data [6\,7] and discuss the implica
 tions of such architectures for generalization of generative models. The w
 orkshop\, ultimately\, will serve to disseminate recent advances across co
 mmunities and reflect on the next steps to tackle several critical challen
 ges in the field.\n\nReferences\n[1] F. Noé\, S. Olsson\, J. Köhler\, H.
  Wu\, Science\, 365\, (2019)\n[2] Y. Wang\, L. Herron\, P. Tiwary\, Proc.
  Natl. Acad. Sci. U.S.A.\, 119\, (2022)\n[3] M. Albergo\, G. Kanwar\, P. 
 Shanahan\, Phys. Rev. D\, 100\, 034515 (2019)\n[4] M. Gabrié\, G. Rotsko
 ff\, E. Vanden-Eijnden\, Proc. Natl. Acad. Sci. U.S.A.\, 119\, (2022)\n[5
 ] S. Ciarella\, J. Trinquier\, M. Weigt\, F. Zamponi\, Mach. Learn.: Sci. 
 Technol.\, 4\, 010501 (2023)\n[6] S. Cheng\, L. Wang\, T. Xiang\, P. Zhan
 g\, Phys. Rev. B\, 99\, 155131 (2019)\n[7] J. Nigam\, S. Pozdnyakov\, G. 
 Fraux\, M. Ceriotti\, The Journal of Chemical Physics\, 156\, (2022)
LOCATION:BCH 2103 https://plan.epfl.ch/?room==BCH%202103
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR