AlphaProof

10 statements across 1 episodes · 6 bullish · 0 bearish · 4 people on the record · first statement Nov 14, 2024 by Rishi Mehta · said 27 times in 1 episodes since 2024 · across every show →

Mentions by year

brought up most by Sarah Guo (17), Elad Gil (6), Laurent Sartran (3), Thomas Hubert (1)

tap a year for its mentions
001513012024episodesmentions
0112024episodes it came up in
00150.53012024episodesmentions per episode
2024 27 mentions in 1 episode

every mention, scene by scene, with the transcript →

Everything said about AlphaProof, oldest first

Nov 14, 2024 neutral
Assertion Not checkable as stated
Mehta: Fields Medalist Tim Gowers failed to find AlphaProof's IMO construction
“Tim Gowers, who was one of our judges and is also a fields medalist. Tried this question for a couple hours and he couldn't find the construction for function that had this property”
Rishi Mehta Nov 14, 2024 ▶ 33:55 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 neutral
Assertion Supported
Hubert: AlphaProof reached high school level but cannot rival Terence Tao
“And to be honest with you know, like kind of we, at the moment we can't rival at all with someone like Terry Tao. We, I think we demonstrated that what we've demonstrated is that we can learn general mathematics almost from scratch and arrive at kind of an imp…”
Thomas Hubert Nov 14, 2024 ▶ 28:36 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 bullish
Opinion
Mehta: AlphaProof's RL scaling and test-time compute generalize across domains
“Some of the sort of tech we developed here of like, you know, like scaling RL and like figuring out how to spend a lot of inference time compute stuff like this feels like it's Quite generally applicable to many other problems.”
Rishi Mehta Nov 14, 2024 ▶ 21:49 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 neutral
Disclosure
Sartran: AlphaProof develops an alien style by learning from self-discovered proofs
“The way alpha proof currency operates there is that it discovers its own proofs, and when they are valid, it learns from them and develop its own style which has been commented upon as looking, yeah, quite, quite alien.”
Laurent Sartran Nov 14, 2024 ▶ 24:57 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 neutral
Assertion Supported
Sartran: AlphaProof cannot perform theory building or invent new theories
“Maybe the main thing that Alphaproof doesn't do is theory building. It doesn't invent its own theories.”
Laurent Sartran Nov 14, 2024 ▶ 10:04 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 positive
Assertion Supported
Mehta: AlphaProof Is Strongest in IMO Algebra and Number Theory
“So the IMO problems have come in four categories. So there's algebra number theory, combinatrix, and geometry. The two that it's strongest at are algebra and number theory. It's relatively weaker at combinatrix, although it can do quite good at some IMO combin…”
Rishi Mehta Nov 14, 2024 ▶ 7:51 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 bullish
Assertion Not checkable as stated
Sartran: Translating all known human proofs would save years of compute
“With more supervised data, we can avoid the exploration problem, and we could translate all the proofs that are known to man, and that certainly would make the agent much better, and that would save us years of compute for sure.”
Laurent Sartran Nov 14, 2024 ▶ 26:07 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 positive
Assertion Partly supported
Guo: DeepMind's AlphaProof solved four of six 2024 IMO problems
“Alphaproof had this really amazing results of solving four of the six problems this year.”
Sarah Guo Nov 14, 2024 ▶ 3:05 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 positive
Insight
Hubert: Formal math proofs enable self-improving reinforcement learning loops
“The advantage of that is that once kind of the proof is complete then you know, the machine would give you a signal back to say, yes, your proof is correct or not. And so we could search for kind of correct proofs. Once we find a correct proof, we can learn fr…”
Thomas Hubert Nov 14, 2024 ▶ 5:55 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Nov 14, 2024 positive
Insight
Mehta: AlphaProof Uses Test-Time RL on Problem Variations to Find Proofs
“One of the ways in which it navigates this massive search space is via an idea that we came up with, which we call test time RL. So this is an idea where, like, let's say you're confronted with a problem that you don't know how to solve. And you know, you can …”
Rishi Mehta Nov 14, 2024 ▶ 8:13 No Priors Ep. 90 | With Google's DeepMind's AlphaProof Team
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 100 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.