10 years ago, ICLR had just 143 submissions. ICLR 2026 received 19,525 valid submissions. Just before the ICLR 2027 deadline, we saw submission ID #62,302. The OpenReview ID isn't the final submission count, but the…

8 points•luxshan_t•11 days ago•6 comments•
10 years ago, ICLR had just 143 submissions. ICLR 2026 received 19,525 valid submissions. Just before the ICLR 2027 deadline, we saw submission ID #62,302. The OpenReview ID isn't the final submission count, but the growth is pretty wild to see.

6 comments

giacomoforte11 days ago
Used to be in every AI/ML job advert, that they preferred those who had previously published in the top 3.

Ironically now it has become a bit of a meme and I see it less often as one of the "nice to have".

sivasuthan911 days ago
This is wild. I submitted at 11:08am IST, the submission ID was already #55,005. By 11:45am IST it had reached #55,499. a HUGE submission number!!!
sivasuthan911 days ago
Really curious to see what research will look like 5–10 years from now :)
jubeerathan11 days ago
1,500,000 submissions for ICLR.
dcminter10 days ago
To save those browsing the google: "International Conference on Learning Representation" so an AI conference.

> ... what happened?

Seriously?

sivasuthan911 days ago
there’s another interesting problem, do authors always understand what they submit?

Some CMU researchers looked into exactly this and proposed a method to tackle it: https://www.cs.cmu.edu/~nihars/preprints/greCAPTCHA.pdf

Read the full thread on Hacker News →

Related stories