BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//compute-paris-2026//speaker//QWKVYT
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:STANDARD
DTSTART:20251125T000000
TZNAME:CET
TZOFFSETFROM:+0100
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260329T030000
RDATE:20270328T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20261025T030000
RDATE:20271031T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:30 replications later: the failure modes nobody reports and how to
  make replications FAIR and citable - Anne fouilloux\, Jean Iaquinta
DTSTART;TZID=Europe/Paris:20261125T151500
DTEND;TZID=Europe/Paris:20261125T154500
DTSTAMP:20260930T112635Z
UID:pretalx-compute-paris-2026-7XUN33@pretalx.com
DESCRIPTION:We don't replicate whole papers — we replicate claims. Over 
 the past year we targeted\nmore than thirty specific published claims acro
 ss ecology\, biodiversity\, and climate\,\nrebuilding each as a self-conta
 ined\, FAIR-by-construction replication you can fork from\nan open templat
 e. This talk shares the failure modes we hit again and again — silent\nd
 ata-substrate bias\, results that hold only under specific priors\, hidden
  train/test\nleakage — and how to catch them. Then it shows how we make 
 each replication count:\nevery finding\, the software behind it\, and even
  the FAIR rules of the domain are\npublished as nanopublications — small
 \, single-claim\, verifiable units — turning one-off\nnotebooks into a c
 itable\, AI-discoverable graph of reproduction evidence you can\nactually 
 trust\, with no hallucinated references.
LOCATION:Room 106
URL:https://pretalx.com/compute-paris-2026/talk/7XUN33/
END:VEVENT
END:VCALENDAR