Schedule
Room 101-B, April 26, 2026
See also the full schedule on the ICLR website.
| Time (BRT) | Event |
|---|---|
| 9:00 - 9:30 | Invited talk 1 - Jeremy Cohen (Flatiron Institute) : How does gradient descent work? |
| 9:30 - 10:00 | Invited talk 2 - Julia Kempe (NYU & Meta) : Some Insights into LLM Reasoning |
| 10:00 - 10:30 | Invited talk 3 - David Bau (Northeastern University) : Reading Science Back Out of AI |
| 10:30 - 11:00 | Coffee Break |
| 11:00 - 11:15 | Contributed Talk 1 - Minhak Song (KAIST) : Zeroth-Order Optimization at the Edge of Stability |
| 11:15 - 11:30 | Contributed Talk 2 - Jingwen Liu (Columbia University) : Less Data, Faster Training: sampling bias from small dataset can speed up training |
| 11:30 - 11:45 | Contributed Talk 3 - Bruno Loureiro (CNRS) : Optimal scaling laws in learning hierarchical multi-index models |
| 11:45 - 12:30 | Poster Session 1 |
| 12:30 - 13:30 | Lunch Break |
| 13:30 - 14:00 | Invited talk 4 - Richard Baraniuk (Rice University & OpenStax) : The science of self-consuming loops in AI |
| 14:00 - 14:30 | Invited talk 5 - Matthieu Wyart (Johns Hopkins University & EPFL): Deriving Neural Scaling Laws from the statistics of natural language |
| 14:30 - 15:00 | Coffee Break |
| 15:00 - 16:00 | Panel Discussion: Jeremy Cohen, Julia Kempe, David Bau, Richard Baraniuk, Matthieu Wyart |
| 16:00 - 16:15 | Challenge Winners Announcement |
| 16:15 - 17:00 | Poster Session 2 |