RSS

The Fu­sion Power Gen­er­a­tor Scenario

johnswentworth
8 Aug 2020 18:31 UTC
42 points
9 comments3 min readLW link

An­a­lyz­ing the Prob­lem GPT-3 is Try­ing to Solve

adamShimi
6 Aug 2020 21:58 UTC
15 points
2 comments4 min readLW link

Mea­sur­ing hard­ware overhang

hippke
5 Aug 2020 19:59 UTC
32 points
4 comments4 min readLW link

[AN #111]: The Cir­cuits hy­pothe­ses for deep learning

rohinmshah
5 Aug 2020 17:40 UTC
23 points
0 comments9 min readLW link
(mailchi.mp)

In­finite Data/​Com­pute Ar­gu­ments in Alignment

johnswentworth
4 Aug 2020 20:21 UTC
41 points
6 comments2 min readLW link

In­ter­pretabil­ity in ML: A Broad Overview

lifelonglearner
4 Aug 2020 19:03 UTC
29 points
0 comments15 min readLW link

Three men­tal images from think­ing about AGI de­bate & corrigibility

steve2152
3 Aug 2020 14:29 UTC
41 points
35 comments4 min readLW link

[Question] What are the most im­por­tant pa­pers/​post/​re­sources to read to un­der­stand more of GPT-3?

adamShimi
2 Aug 2020 20:53 UTC
25 points
3 comments1 min readLW link

In­ner Align­ment: Ex­plain like I’m 12 Edition

Rafael Harth
1 Aug 2020 15:24 UTC
98 points
8 comments12 min readLW link

Power as Easily Ex­ploitable Opportunities

TurnTrout
1 Aug 2020 2:14 UTC
26 points
5 comments6 min readLW link

“Go west, young man!”—Prefer­ences in (im­perfect) maps

Stuart_Armstrong
31 Jul 2020 7:50 UTC
19 points
10 comments3 min readLW link

[Question] What if memes are com­mon in highly ca­pa­ble minds?

Daniel Kokotajlo
30 Jul 2020 20:45 UTC
24 points
5 comments2 min readLW link

Learn­ing the prior and generalization

evhub
29 Jul 2020 22:49 UTC
32 points
10 comments4 min readLW link

What Failure Looks Like: Distill­ing the Discussion

Ben Pace
29 Jul 2020 21:49 UTC
53 points
10 comments7 min readLW link

[AN #110]: Learn­ing fea­tures from hu­man feed­back to en­able re­ward learning

rohinmshah
29 Jul 2020 17:20 UTC
13 points
2 comments10 min readLW link
(mailchi.mp)

The “best pre­dic­tor is mal­i­cious op­ti­miser” problem

Donald Hobson
29 Jul 2020 11:49 UTC
14 points
10 comments2 min readLW link

[Question] What hap­pens to var­i­ance as neu­ral net­work train­ing is scaled? What does it im­ply about “lot­tery tick­ets”?

abramdemski
28 Jul 2020 20:22 UTC
23 points
2 comments1 min readLW link

[Question] To what ex­tent are the scal­ing prop­er­ties of Trans­former net­works ex­cep­tional?

abramdemski
28 Jul 2020 20:06 UTC
25 points
1 comment1 min readLW link

[Question] Does the lot­tery ticket hy­poth­e­sis sug­gest the scal­ing hy­poth­e­sis?

Daniel Kokotajlo
28 Jul 2020 19:52 UTC
13 points
2 comments1 min readLW link

[Question] Prob­a­bil­ity that other ar­chi­tec­tures will scale as well as Trans­form­ers?

Daniel Kokotajlo
28 Jul 2020 19:36 UTC
20 points
4 comments1 min readLW link