TagLast edit: 22 Jul 2022 2:45 UTC by Alex_Altair

A Quantilizer is a proposed AI design which aims to reduce the harms from Goodhart’s law and specification gaming by selecting reasonably effective actions from a distribution of human-like actions, rather than maximizing over actions. It it more of a theoretical tool for exploring ways around these problems than a practical buildable design.

See also

Soft op­ti­miza­tion makes the value tar­get bigger

Jeremy Gillen2 Jan 2023 16:06 UTC
106 points
20 comments12 min readLW link

Another view of quan­tiliz­ers: avoid­ing Good­hart’s Law

jessicata9 Jan 2016 4:02 UTC
26 points
2 comments2 min readLW link

Quan­tiliz­ers max­i­mize ex­pected util­ity sub­ject to a con­ser­va­tive cost constraint

jessicata28 Sep 2015 2:17 UTC
33 points
3 comments5 min readLW link

Quan­tiliz­ers and Gen­er­a­tive Models

Adam Jermyn18 Jul 2022 16:32 UTC
24 points
5 comments4 min readLW link

When to use quantilization

RyanCarey5 Feb 2019 17:17 UTC
65 points
5 comments4 min readLW link

Quan­tilizer ≡ Op­ti­mizer with a Bounded Amount of Output

itaibn016 Nov 2021 1:03 UTC
11 points
4 comments2 min readLW link

[Question] Why don’t quan­tiliz­ers also cut off the up­per end of the dis­tri­bu­tion?

Alex_Altair15 May 2023 1:40 UTC
25 points
2 comments1 min readLW link

Re­cur­sive Quan­tiliz­ers II

abramdemski2 Dec 2020 15:26 UTC
30 points
15 comments13 min readLW link

Stable Poin­t­ers to Value III: Re­cur­sive Quantilization

abramdemski21 Jul 2018 8:06 UTC
19 points
4 comments4 min readLW link
No comments.