NeurIPS 2025 Retrospective
Here is the post series for the conference.
NeurIPS 2025 is upon us! Expo talks and tutorials started December 2nd. I am fortunate that my work provides a pretty large learning budget each year which can be used on conferences. I attended PyCon 2025 this year, which took up a portion of the budget, so I did not attend any tutorials on the first day of neurIPS. On the plus side, one less day of a conference means I won't be as exhausted. [1]
This year the conference is in both San Diego and Mexico City. Partially because of how many people want to attend, but also because of visa accessibility problems in the US. I am attending at the San Diego location. The full schedule is available online. [2]
Unlike PyCon, I don't know if the recordings will be made available online for non-attendees. Some of the links I provide in the section titles include slides though.
Also, sometimes the descriptions will be a little terse. I promise you, I wrote them myself, I'm was just out of brain juice after the long days.
I have notes from the three days I attended:
Why did I attend NeurIPS and a thank you to my coworkers
I work on software products that use machine learning under the hood. However, I am squarely on the applications side in the org chart. I get to work closely with the ML side though and I enjoy that a lot. Last year I went to EMNLP 2024 in Miami with the goal of reducing my unknown unknowns. Ultimately, I want to understand the different choices that go into the ML side so I can better collaborate on projects.
At EMNLP I was able to spend time with the ML folks and learn from them. For that I am really appreciative. They showed me the ropes for attending an academic conference like that. Some of my takeaways were:
- Take notes! There will be so much content you will not remember all of it.
- Pace yourself. Its tempting to go to everything, but that's not sustainable.
- On the flip side, you get out of a conference what you put into it. If you just randomly go to things you won't gain much.
This time at NeurIPS I am leaning on them the same way. Some new lessons:
- If you're not sure what talks to go to, have a look at the best paper awards and use that as a starting point.
- If you want the coveted conference branded water bottle, go early because they are gone quickly.
- For some reason now, there are cafes in the expo hall. The companies fight over hiring the best baristas they can. [3]
- My coworkers also went to a event by Weights & Biases after the sessions. I didn't sign up in time so I couldn't go, but this was recommended because we already use their product at work. That means they won't bother selling you anything and you can eat the free food unbothered.
The Cafes:
Also, somewhat uncharacteristically, I organized the team dinner this time.
For some reason, people are writing a lot of papers about AI right now
This is my first time attending NeurIPS, so I cannot compare my experience to previous years, but it seems there were a lot of papers this time. Melanie Mitchell mentioned in her talk that in the 90s, NeurIPS was just a couple hundred people and they would spend most of the day skiing. In comparison this year is the 39th conference and in the conference's reflections post they mention "The main track this year received 21,575 valid paper submissions, of which 5,290 were accepted". I heard some people mention that papers were accepted, but then later rejected due to limited space. I can't find a solid source to cite though.
The San Diego Convention Center was huge. I would frequently underestimate how much time I needed to get from one side to another to catch a talk. The fact that they still didn't have enough space really speaks to just how many papers were submitted.
Finally a bit of meta commentary on writing
I've been trying to practice new kinds of writing. At this conference I wanted to challenge myself to try to write and share summary posts as close to daily as possible. The process felt a little like journalism, though I never took a class on that at school. I'm not sure if I like it, but the stretch was a nice challenge.
The downside was that I sometimes felt like I was missing out on the conference since my attention was split with writing. I think in the future I will take notes, but not edit and post them until after the conference is over.
Also Meshtastic
Totally unrelated to all of this, but I brought my MeshPocket so I could see the San Diego and Californian Meshnets when I was in town. It's looking pretty healthy:
I managed to get one node over the boarder in Mexico, and not pictured here, but I hopped to some nodes in LA.
No actually, no matter what, you are tired after a conference like this. Especially with a 3 hour time change. ↩︎
I did not review the schedule ahead of time, and boy do I wish I did. The way it was presented made it very hard to parse. ↩︎
Certainly there is no bubble, and this is not a gilded age. ↩︎
Member discussion