OpenAI claims gold-medal performance at IMO 2025

Waiting for Terry Tao's thoughts, but these kind of things are good use of AI. We need to make science progress faster rather than disrupting our economy without being ready.

Re: OpenAI claims gold-medal performance at IMO 2025

#5

Post by ktallett » Sat, Jul 19, 2025, 10:25 AM UTC

[flagged]

Re: OpenAI claims gold-medal performance at IMO 2025

#6

Post by Lionga » Sat, Jul 19, 2025, 10:26 AM UTC

[flagged]

Re: OpenAI claims gold-medal performance at IMO 2025

#7

Post by ktallett » Sat, Jul 19, 2025, 10:27 AM UTC

demirbey05 wrote: post #4

Progress is astounding. Recently report published about evaluation of LLMs on IMO 2025. o3 high didn't even get bronze. https://matharena.ai/imo/ Waiting for Terry Tao's thoughts, but these kind of things are good use of AI. We need to make science progress faster rather than disrupting our economy without being ready.

[flagged]

Re: OpenAI claims gold-medal performance at IMO 2025

#8

Post by ktallett » Sat, Jul 19, 2025, 10:28 AM UTC

Lionga wrote: post #6

[flagged]

[flagged]

Re: OpenAI claims gold-medal performance at IMO 2025

#9

Post by z7 » Sat, Jul 19, 2025, 10:34 AM UTC

Some previous predictions:

In 2021 Paul Christiano wrote he would update from 30% to "50% chance of hard takeoff" if we saw an IMO gold by 2025.

He thought there was an 8% chance of this happening.

Eliezer Yudkowsky said "at least 16%".

Source:

https://www.lesswrong.com/posts/sWLLdG6DWJEy3CH7n/imo-challe...

Re: OpenAI claims gold-medal performance at IMO 2025

#10

Post by littlestymaar » Sat, Jul 19, 2025, 10:37 AM UTC

ktallett wrote: post #3

tester756 wrote: post #2

huh? any details?

[flagged]

Which would be impressive if we knew those problems weren't in the training data already.

I mean it is quite impressive how language models are able to mobilize the knowledge they have been trained on, especially since they are able to retrieve information from sources that may be formatted very differently, with completely different problem statement sentences, different variable names and so on, and really operate at the conceptual level.

But we must wary of mixing up smart information retrieval with reasoning.