Epoch confirms GPT5.4 Pro solved a frontier math open problem
1–10 of 744 posts
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#2> Subsequent to this solve, we finished developing our general scaffold for testing models on FrontierMath: Open Problems. In this scaffold, several other models were able to solve the problem as well: Opus 4.6 (max), Gemini 3.1 Pro, and GPT-5.4 (xhigh).
Interesting. Whats that “scaffold”? A sort of unit test framework for proofs?
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#3> Subsequent to this solve, we finished developing our general scaffold for testing models on FrontierMath: Open Problems. In this scaffold, several other models were able to solve the problem as well: Opus 4.6 (max), Gemini 3.1 Pro, and GPT-5.4 (xhigh). Interesting. Whats that “scaffold”? A sort of unit test framework for proofs?
I think in this context, scaffolds are generally the harness that surrounds the actual model. For example, any tools, ways to lay out tasks, or auto-critiquing methods.
I think there's quite a bit of variance in model performance depending on the scaffold so comparisons are always a bit murky.
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#4No denial at this point, AI could produce something novel, and they will be doing more of this moving forward.
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#5Seems like the high compute parallel thinking models weren't even needed, both the normal 5.4 and gemini 3.1 pro solved it. Somehow Gemini 3 deepthink couldn't solve it.
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#6Fantastic news! That means with the right support tooling existing models are already capable of solving novel mathematics. There’s probably a lot of good mathematics out there we are going to make progress on.
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#7No denial at this point, AI could produce something novel, and they will be doing more of this moving forward.
[flagged]
Re: Epoch confirms GPT5.4 Pro solved a frontier math open problem
#8> Subsequent to this solve, we finished developing our general scaffold for testing models on FrontierMath: Open Problems. In this scaffold, several other models were able to solve the problem as well: Opus 4.6 (max), Gemini 3.1 Pro, and GPT-5.4 (xhigh). Interesting. Whats that “scaffold”? A sort of unit test framework for proofs?
I think in this context, scaffolds are generally the harness that surrounds the actual model. For example, any tools, ways to lay out tasks, or auto-critiquing methods. I think there's quite a bit of variance in model performance depending on the scaffold so comparisons are always a bit murky.
Usually involves a lot of agents and their custom contexts or system prompts.