Aaron should hopefully know. Last year was it 15 minutes + 5 minutes questions? or 20 minutes + 5 minutes questions. Something like this.
However I would do a single practice talk per session of tea talks, because from experience, we usually end up analysing each slide and figure out how to improve it .. which ends up taking quite some time.
Razvan
On Thu, Mar 13, 2014 at 11:18 AM, Ian Goodfellow goodfellow.ian@gmail.comwrote:
Do we know how long the oral presentations at ICLR should be?
2014-03-13 11:09 GMT-04:00 Razvan Pascanu r.pascanu@gmail.com:
Hi guys,
First of all, those with accepted papers at ICLR.
I know I and Ian have an oral. Does anybody else got his paper accepted as an oral. If so and you want to do a practice talk let me know to reserve a spot. Ian, do you think you need a practice talk?
Otherwise, those with posters, if you think having a tea talk will help you talk about the poster, or if you want to present your work, let me know so that I can schedule you.
For now we have: next week (19th) is me + Yann talking about non-convex optimizations and saddle points 26th we have Mehdi talking about his paper at ICLR (catastrophic forgetting). 2nd of April I scheduled myself to do a practice talk for my oral (natural gradient)
Mehdi said his talk is going to be short, so he would not mind if someone else wants to slit that time slot with him. Anybody wants to do that ?
If no one with ICLR paper wants to present their work, is there someone else who wants to present either:
(a) an idea they are working on (b) some submitted work somewhere (or just tech report) (c) some very interesting paper they read recently (d) just introduce themselves to the lab talking of what they've done before coming here
Thanks a lot, Best, Razvan
Lisa_labo mailing list Lisa_labo@iro.umontreal.ca https://webmail.iro.umontreal.ca/mailman/listinfo/lisa_labo