High Signal Podcasts Evidence ledger
Method
Browse
How the evidence works

Public evidence record

Kenneth Stanley

Published podcast speaker

Claims
14
Episodes
2
Shows
1
Named items
0

Claim ledger

What Kenneth said.

4 transcript-backed records

01 / evaluation

We also show new examples. But you you know, the word reason I use the word indirect is because unlike these pick breeder images, we can't just go in and look at a neuron and know what it does explicitly because that's what's so nice about pick breeder images because they're 2 d.

“We also show new examples. But you you know, the word reason I use the word indirect is because unlike these pick breeder images, we can't just go in and look at a neuron and know what it does explicitly because that's what's so nice about pick breeder images because they're 2 d.”
Speaker
Kenneth Stanley
Publisher
Machine Learning Street Talk

02 / evaluation

It's just emergent from how SGD climbs these gradients. But the thing that I think makes this really intriguing, the reason that the paper is because beyond just that, is that it gives you something that otherwise could never exist, which is a counterexample, that there actually do exist networks that don't have that issue.

“It's just emergent from how SGD climbs these gradients. But the thing that I think makes this really intriguing, the reason that the paper is because beyond just that, is that it gives you something that otherwise could never exist, which is a counterexample, that there actually do exist networks that don't have that issue.”
Speaker
Kenneth Stanley
Publisher
Machine Learning Street Talk

03 / evaluation

Like, what are you objecting to? But they but the point is that it can still be an impostor because it's like, what we care about here is not just that it's going to get answers right, like, good test scores, like, seem to be plausibly human when you talk about things that are in distribution.

“Like, what are you objecting to? But they but the point is that it can still be an impostor because it's like, what we care about here is not just that it's going to get answers right, like, good test scores, like, seem to be plausibly human when you talk about things that are in distribution.”
Speaker
Kenneth Stanley
Publisher
Machine Learning Street Talk

04 / evaluation

The thing that I think makes this really intriguing is that it gives you something that otherwise could never exist, which is a counterexample that there actually do exist networks that don't have that issue.

“The thing that I think makes this really intriguing is that it gives you something that otherwise could never exist, which is a counterexample that there actually do exist networks that don't have that issue.”
Speaker
Kenneth Stanley
Publisher
Machine Learning Street Talk
Search evidence