As a Math PhD student who uses LLMs for math every day, I can say that there is still a long way to go before LLMs can effectively solve these problems. The only scenario where I see this happening soon is if models like AlphaProof scale significantly more than expected within the next year.

@Grothenfla have you tried using our lord and savior o1-pro ?

@MalachiteEagle Not yet... But o1-pro's AIME benchmark is similar to this new model Kimina (86%). I believe these formal theorem provers are the real contenders.

@Grothenfla i'm curious what you think of the new natural language models that got gold on the IMO? Like has your view updated? I think that also alphaproof-like systems will become quite good but it's less clear to me than it was when you wrote that comment that they would be better than the general reasoners

@Bayesian Yes, I have updated my view after Gemini's and OpenAI. To be honest I thought that we were very far away from a achievement like this. We still know little about these LLMs but now I think it is possible that they become as good as the best mathematicians in the coming years. But I still think systems like alphaproof still have a better chance in becoming superhuman mathematicians similar to AlphaGo and AlphaZero becoming superhuman in their respective areas.

@Grothenfla yeah i pretty much agree! the only consideration that goes against that is that i could imagine there being like several orders of magnitude (200x?) more research effort and compute allocated to general reasoners than to alphaproof/alphazero-like systems, and there could be enough far transfer or wtv, that this ends up being wrong? Ig i'd add that the narrow math systems would likely become much better at proofs for a similar amt of effort, but probably are hopeless for some tasks mathematicians have to do that are more diverse than theorem proving, like deciding which research direction to go into next and so on. exciting times..?

o4 is coming

@MalachiteEagle by end of year we might already have o6 (build into gpt5 or whatever they will call it)

@sponge can you clarify whether you will use official epoch results or outside claims?

fk it i'll buy more at 35%

@Bayesian Did something change your mind? It seems like Grok 3 isn't quite at the level of o3 on math and coding benchmarks, even with reasoning enabled.

@TimothyJohnson5c16 polymarket has a market that has given me confidence that I am right

https://polymarket.com/event/ai-model-scores-90-on-frontiermath-benchmark-in-2025?tid=1739933819530

(90% is much harder than 80% bc there's around a 10% error rate, ie ~10% of problems can't be solved correctly according to the benchmark)

@Bayesian Yeah, if I were on Polymarket, I would just call that free money.

@Bayesian Though of course, 10% error rate is also an estimate, right? In the world where an AI model reaches 90%, the error rate is probably a lot lower.

@TimothyJohnson5c16 yeah 10% error is an estimate, the errors they found were like 6% iirc, and they estimated # of errors they would have missed

@Bayesian that a big deal if there's a 10% error rate though right?

@NebulaByte I don’t understand your comment could you rephrase

@Bayesian what did you mean by 10% error in your comment above?

@NebulaByte i think i mean that about 10% of (question, answer) pairs in the benchmark are eroneous in some way, either that the question has some hidden assumptions that make it not necessarily solvable, or the answer is incorrect (the person solving the problem to add it to the benchmark made a mistake in some step) or something like that

@Bayesian want to buy more? I put a limit order at 35%.

Lol, Acceleration took most of it, so I added some more.

@TimothyJohnson5c16 I was scared but i change my mind ill buy more. YES order at 36%

@Bayesian if you still want more I put a limit order at 40

@Bayesian I think yours was taken already? I put more NO at 35%.