Gary Marcus

Rank 12 of 47
|
Score 83
5/19/2025, 12:36:32 AM
X
In reply to:
Daniel Litt
@littmath
·
251d
@GaryMarcus @TimHenke9 @Nicky_Bonez Ok, I think it’s quite strong evidence. I agree it’s not definitive on its own but it’s not as if it’s the only example we have of high performance OOD.
Gary Marcus
@GaryMarcus
·
251d
@littmath @TimHenke9 @Nicky_Bonez It’s *evidence*, but not *definitive* for all the reasons I have enumerated in our recent discussion.
Daniel Litt
@littmath
·
251d
@GaryMarcus @TimHenke9 @Nicky_Bonez Ok, if we agree on this, it would help if you say why you don’t think high performance on OOD math questions isn’t evidence for intelligence.
Gary Marcus
@GaryMarcus
·
251d
@littmath @TimHenke9 @Nicky_Bonez ps, put differently, we roughly agree on the facts for the math questions, but disagree on what lessons you can draw from there.
Daniel Litt
@littmath
·
251d
@GaryMarcus @TimHenke9 @Nicky_Bonez FWIW I’m quite confident the questions I’m asking are OOD. I agree it only solves some; unclear to me why this matters.
Gary Marcus
@GaryMarcus
·
251d
@littmath @TimHenke9 @Nicky_Bonez they can solve SOME hard problems

BUT you are ignoring a lot.

1. we don’t know which math problems are OOD, because we don’t know what is in the training set and augmented data

2. we do know that performance on the USAMO, which tried to preclude data contamination, was far

The statement is part of a public discourse as it engages in a substantive discussion about the interpretation of evidence related to high performance on out-of-distribution (OOD) math questions, which is a topic of public interest in the field of artificial intelligence and education. The tone of the statement is one of frustration, as the speaker feels their arguments are being ignored, which they find disrespectful.

  1. Principle 4:
    I will engage in constructive criticism and dialogue with those in disagreement and will not engage in personal attacks or ad hominem arguments.
    The statement expresses frustration about not being engaged in arguments, which could be seen as a call for more constructive dialogue. However, it does not engage in personal attacks or ad hominem arguments, thus adhering to the principle of engaging in constructive criticism and dialogue. [+1]