BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//ox.ac.uk//NONSGML oxford.event//EN
X-WR-TIMEZONE:Europe/London
BEGIN:VTIMEZONE
TZID:Europe/London
X-LIC-LOCATION:Europe/London
BEGIN:DAYLIGHT
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:BST
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:GMT
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Joint OxfordXML seminar together with the 6th Signal Processing an
 d Monitoring (SPaM) in Labour International Workshop to this set of talks 
 on machine learning
DTSTART;TZID=Europe/London:20260625T154500
DTEND;TZID=Europe/London:20260625T181500
DTSTAMP:20260619T054541Z
UID:080743b8-976a-f111-ab0d-7c1e52046959
CREATED:20260617T215850Z
DESCRIPTION:The Oxford XML Cluster is pleased to invite you to a joint sem
 inar together with the 6th Signal Processing and Monitoring (SPaM) in Labo
 ur International Workshop to this set of talks on machine learning\, appli
 cations and more.\n\nDate: Thu\, 25 Jun 2026 | 15:45 - 18:15 (come and go 
 as you please)\nLocation: Wolfson College\, Leonard Wolfson Auditorium\nEv
 ent URLs: OxfordXML: The Cross-disciplinary Machine Learning Community (ht
 tps://oxfordxml.github.io/) and SPaM in Labour Workshop (https://users.ox.
 ac.uk/~ndog0178/spam2026.htm)\nMS Teams Event: https://teams.microsoft.com
 /l/meetup-join/19%3ameeting_NzA1OWUzMTUtY2ViYS00YTJmLTk2YTYtOGMwZTY3N2IyMj
 Nl%40thread.v2/0?context=%7b%22Tid%22%3a%22cc95de1b-97f5-4f93-b4ba-fe68b85
 2cf91%22%2c%22Oid%22%3a%222d6d82c4-6b2c-4f77-b979-7c49923c3b36%22%7d\n\nSc
 hedule:\n\nCake and coffee in the Buttery at 15:45-16:15\n\n16:15-16:35 Sh
 eng Wong\, University of Oxford (UK)\nPRISM-CTG: A Foundation model for ca
 rdiotocography analysis with multi-view SSL  \n\n16:35-16:55 Maria Signori
 ni\, University Milano (Italy)\nAnalysis of Fetal Heart Rate antepartum: m
 ultiparametric methods and artificial\nintelligence contribution\n\n16:55-
 17:15 Martin Frasch (US)\nPregnancy health monitoring: where are we headed
 ? Experiences using the recently\nreleased 10M ECG foundation model and mo
 re\n\n17:15-18:00 XML Cluster event: Peter Koepernik\, OpenAI\nMemory Lear
 ning under Partial Observability\n\nOpen end: Whole room discussion\n\nAbs
 tract for Dr Koepernik's talk: When a reinforcement learning agent has acc
 ess only to partial observations of its environment\, optimal decision-mak
 ing generally requires retaining and using information from the past. This
  work characterizes the properties a learned memory representation must sa
 tisfy for an optimal policy to be expressible as a function of that repres
 entation. Building on this\, we introduce an auxiliary training objective 
 that encourages deep reinforcement learning agents to learn such memory fu
 nctions. Empirical results across a diverse set of environments demonstrat
 e that this approach can substantially improve performance under partial o
 bservability.\n\nBio for Dr Koepernik: Peter is a Research Scientist at Op
 enAI working on sub-quadratic attention mechanisms to improve long-context
  performance of large language models. He recently completed a DPhil in St
 atistics at Oxford\, with research in probability theory\, stochastic anal
 ysis\, numerical SDE methods\, and reinforcement learning under partial ob
 servability. More broadly\, he is interested in how mathematical approache
 s can help make machine learning algorithms more scalable\, robust\, and u
 seful.
LAST-MODIFIED:20260617T221546Z
LOCATION:Wolfson College - Leonard Wolfson Auditorium\, Leonard Wolfson Au
 ditorium Wolfson College Linton Road Oxford Oxfordshire OX2 6UD United Kin
 gdom
END:VEVENT
END:VCALENDAR