All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 202320242025

All Jan Feb Mar Apr MayJunJul Aug Sep Oct Nov Dec

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 171819 20 21 22 23 24 25 26 27 28 29 30

D&D.Sci Alchemy: Archmage Anachronos and the Supply Chain Issues Evaluation & Ruleset

aphyerJun 17, 2024, 9:29 PM

61 points

17 votes

Overall karma indicates overall quality.

11 comments6 min readLW link

Questionable Narratives of “Situational Awareness”

fergusqJun 17, 2024, 9:01 PM

0 points

5 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

(forum.effectivealtruism.org)

ZuVillage Georgia – Mission Statement

BurnsJun 17, 2024, 7:53 PM

3 points

13 votes

Overall karma indicates overall quality.

3 comments9 min readLW link

Getting 50% (SoTA) on ARC-AGI with GPT-4o

ryan_greenblattJun 17, 2024, 6:44 PM

263 points

127 votes

Overall karma indicates overall quality.

50 comments13 min readLW link

Sycophancy to subterfuge: Investigating reward tampering in large language models

Carson Denison and evhub

Jun 17, 2024, 6:41 PM

163 points

54 votes

Overall karma indicates overall quality.

22 comments8 min readLW link

(arxiv.org)

Labor Participation is a High-Priority AI Alignment Risk

alexJun 17, 2024, 6:09 PM

7 points

6 votes

Overall karma indicates overall quality.

0 comments17 min readLW link

Towards a Less Bullshit Model of Semantics

johnswentworth and David Lorell

Jun 17, 2024, 3:51 PM

94 points

37 votes

Overall karma indicates overall quality.

44 comments21 min readLW link

Analysing Adversarial Attacks with Linear Probing

Yoann Poupart, Imene Kerboua, Clement Neo and Jason Hoelscher-Obermaier

Jun 17, 2024, 2:16 PM

9 points

5 votes

Overall karma indicates overall quality.

0 comments8 min readLW link

What’s the future of AI hardware?

Itay DreyfusJun 17, 2024, 1:05 PM

2 points

3 votes

Overall karma indicates overall quality.

0 comments8 min readLW link

(productidentity.co)

OpenAI #8: The Right to Warn

ZviJun 17, 2024, 12:00 PM

97 points

38 votes

Overall karma indicates overall quality.

8 comments34 min readLW link

(thezvi.wordpress.com)

Logit Prisms: Decomposing Transformer Outputs for Mechanistic Interpretability

ntt123Jun 17, 2024, 11:46 AM

5 points

5 votes

Overall karma indicates overall quality.

4 comments6 min readLW link

(neuralblog.github.io)

Weak AGIs Kill Us First

yrimonJun 17, 2024, 11:13 AM

15 points

9 votes

Overall karma indicates overall quality.

4 comments9 min readLW link

[Linkpost] Guardian article covering Lightcone Infrastructure, Manifest and CFAR ties to FTX

ROMJun 17, 2024, 10:05 AM

8 points

9 votes

Overall karma indicates overall quality.

9 comments1 min readLW link

(www.theguardian.com)

Fat Tails Discourage Compromise

niplavJun 17, 2024, 9:39 AM

53 points

27 votes

Overall karma indicates overall quality.

5 comments1 min readLW link

Our Intuitions About The Criminal Justice System Are Screwed Up

Bentham's BulldogJun 17, 2024, 6:22 AM

12 points

11 votes

Overall karma indicates overall quality.

15 comments4 min readLW link

A Case for Cooperation: Dependence in the Prisoner’s Dilemma

grantstengerJun 17, 2024, 1:10 AM

10 points

8 votes

Overall karma indicates overall quality.

3 comments23 min readLW link

Degeneracies are sticky for SGD

Guillaume Corlouer and Nicolas Macé

Jun 16, 2024, 9:19 PM

56 points

25 votes

Overall karma indicates overall quality.

1 comment16 min readLW link

YM’s Shortform

YMJun 16, 2024, 8:57 PM

3 points

2 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

“Is-Ought” is Fraught

MiSteR KitttyJun 16, 2024, 5:27 PM

−5 points

6 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

The type of AI humanity has chosen to create so far is unsafe, for soft social reasons and not technical ones.

l8cJun 16, 2024, 1:31 PM

−6 points

7 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

Self-Control of LLM Behaviors by Compressing Suffix Gradient into Prefix Controller

Henry CaiJun 16, 2024, 1:01 PM

7 points

7 votes

Overall karma indicates overall quality.

0 comments7 min readLW link

(arxiv.org)

CIV: a story

Richard_NgoJun 15, 2024, 10:36 PM

99 points

56 votes

Overall karma indicates overall quality.

6 comments9 min readLW link

(www.narrativeark.xyz)

Yann LeCun: We only design machines that minimize costs [therefore they are safe]

tailcalledJun 15, 2024, 5:25 PM

19 points

13 votes

Overall karma indicates overall quality.

8 comments1 min readLW link

(twitter.com)

(Appetitive, Consummatory) ≈ (RL, reflex)

Steven ByrnesJun 15, 2024, 3:57 PM

38 points

14 votes

Overall karma indicates overall quality.

1 comment3 min readLW link

Two LessWrong speed friending experiments

mikko and sanyer

Jun 15, 2024, 10:52 AM

52 points

38 votes

Overall karma indicates overall quality.

3 comments4 min readLW link

Claude’s dark spiritual AI futurism

jessicataJun 15, 2024, 12:57 AM

22 points

12 votes

Overall karma indicates overall quality.

7 comments43 min readLW link

(unstableontology.com)

[Question] When is “unfalsifiable implies false” incorrect?

VojtaKovarikJun 15, 2024, 12:28 AM

3 points

9 votes

Overall karma indicates overall quality.

11 comments1 min readLW link

MIRI’s June 2024 Newsletter

HarlanJun 14, 2024, 11:02 PM

74 points

27 votes

Overall karma indicates overall quality.

20 comments2 min readLW link

(intelligence.org)

Language for Goal Misgeneralization: Some Formalisms from my MSc Thesis

GiulioJun 14, 2024, 7:35 PM

10 points

6 votes

Overall karma indicates overall quality.

0 comments8 min readLW link

(www.giuliostarace.com)

Shard Theory—is it true for humans?

RishikaJun 14, 2024, 7:21 PM

71 points

32 votes

Overall karma indicates overall quality.

7 comments15 min readLW link

When fine-tuning fails to elicit GPT-3.5′s chess abilities

Theodore ChapmanJun 14, 2024, 6:50 PM

42 points

24 votes

Overall karma indicates overall quality.

3 comments9 min readLW link

Results from the AI x Democracy Research Sprint

Esben Kran, jordine and Jason Hoelscher-Obermaier

Jun 14, 2024, 4:40 PM

13 points

7 votes

Overall karma indicates overall quality.

0 comments6 min readLW link

Rational Animations’ intro to mechanistic interpretability

WriterJun 14, 2024, 4:10 PM

45 points

20 votes

Overall karma indicates overall quality.

1 comment11 min readLW link

(youtu.be)

Why keep a diary, and why wish for large language models

DanielFilanJun 14, 2024, 4:10 PM

9 points

7 votes

Overall karma indicates overall quality.

1 comment2 min readLW link

(danielfilan.com)

The Leopold Model: Analysis and Reactions

ZviJun 14, 2024, 3:10 PM

109 points

30 votes

Overall karma indicates overall quality.

19 comments57 min readLW link

(thezvi.wordpress.com)

[Question] Thoughts on Francois Chollet’s belief that LLMs are far away from AGI?

O OJun 14, 2024, 6:32 AM

26 points

12 votes

Overall karma indicates overall quality.

17 comments1 min readLW link

Research Report: Alternative sparsity methods for sparse autoencoders with OthelloGPT.

Andrew QuaisleyJun 14, 2024, 12:57 AM

17 points

9 votes

Overall karma indicates overall quality.

5 comments12 min readLW link

Slowed ASI—a possible technical strategy for alignment

Lester LeongJun 14, 2024, 12:57 AM

5 points

8 votes

Overall karma indicates overall quality.

2 comments3 min readLW link

Conceptual Typography Example

milanroskoJun 14, 2024, 12:39 AM

15 points

15 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Safety isn’t safety without a social model (or: dispelling the myth of per se technical safety)

Andrew_CritchJun 14, 2024, 12:16 AM

357 points

170 votes

Overall karma indicates overall quality.

38 comments4 min readLW link

OpenAI appoints Retired U.S. Army General Paul M. Nakasone to Board of Directors

Joel BurgetJun 13, 2024, 9:28 PM

35 points

20 votes

Overall karma indicates overall quality.

10 comments1 min readLW link

(openai.com)

AI #68: Remarkably Reasonable Reactions

ZviJun 13, 2024, 4:30 PM

46 points

27 votes

Overall karma indicates overall quality.

11 comments50 min readLW link

(thezvi.wordpress.com)

Four Futures For Cognitive Labor

Maxwell TabarrokJun 13, 2024, 12:56 PM

14 points

11 votes

Overall karma indicates overall quality.

11 comments4 min readLW link

(www.maximum-progress.com)

Underrated Proverbs

Arjun PanicksseryJun 13, 2024, 12:30 PM

13 points

14 votes

Overall karma indicates overall quality.

9 comments1 min readLW link

(arjunpanickssery.substack.com)

[Paper] AI Sandbagging: Language Models can Strategically Underperform on Evaluations

Teun van der Weij, Felix Hofstätter, Ollie J, Sam F. Brown and Francis Rhys Ward

Jun 13, 2024, 10:04 AM

84 points

35 votes

Overall karma indicates overall quality.

10 comments2 min readLW link

(arxiv.org)

Probably Not a Ghost Story

George IngebretsenJun 12, 2024, 10:55 PM

27 points

13 votes

Overall karma indicates overall quality.

4 comments3 min readLW link

AiPhone

ZviJun 12, 2024, 10:20 PM

63 points

23 votes

Overall karma indicates overall quality.

4 comments14 min readLW link

(thezvi.wordpress.com)

microwave drilling is impractical

bhauthJun 12, 2024, 10:16 PM

59 points

32 votes

Overall karma indicates overall quality.

19 comments4 min readLW link

(www.bhauth.com)

Phonosemantic Duplication

bitcoinssgJun 12, 2024, 8:19 PM

5 points

3 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

My AI Model Delta Compared To Christiano

johnswentworthJun 12, 2024, 6:19 PM

191 points

92 votes

Overall karma indicates overall quality.

74 comments4 min readLW link

Keyboard shortcuts

Keys shown in yellow (e.g., ]) are accesskeys, and require a browser-specific modifier key (or keys).

Keys shown in grey (e.g., ?) do not require any modifier keys.

General
? Show keyboard shortcuts
Esc Hide keyboard shortcuts

Site navigation
h Go to Home (a.k.a. “Frontpage”) view
f Go to Featured (a.k.a. “Curated”) view
a Go to All (a.k.a. “Community”) view
m Go to Meta view
v Go to Tags view
c Go to Recent Comments view
r Go to Archive view
q Go to Sequences view
t Go to About page
u Go to User or Login page
o Go to Inbox page

Page navigation
, Jump up to top of page
. Jump down to bottom of page
/ Jump to top of comments section
s Search

Page actions
n New post or comment
e Edit current post

Post/comment list views
. Focus next entry in list
, Focus previous entry in list
; Cycle between links in focused entry
Enter Go to currently focused entry
Esc Unfocus currently focused entry
] Go to next page
[ Go to previous page
\ Go to first page
e Edit currently focused post

Editor
k Bold text
i Italic text
l Insert hyperlink
q Blockquote text

Appearance
= Increase text size
- Decrease text size
0 Reset to default text size
′ Cycle through content width settings
1 Switch to default theme [A]
2 Switch to dark theme [B]
3 Switch to grey theme [C]
4 Switch to ultramodern theme [D]
5 Switch to simple theme [E]
6 Switch to brutalist theme [F]
7 Switch to ReadTheSequences theme [G]
8 Switch to classic Less Wrong theme [H]
9 Switch to modern Less Wrong theme [I]
; Open theme tweaker
Enter Save changes and close theme tweaker
Esc Close theme tweaker (without saving)

Slide shows
l Start/resume slideshow
Esc Exit slideshow
→↓ Next slide
←↑ Previous slide
Space Reset slide zoom

Miscellaneous
x Switch to next view on user page
z Switch to previous view on user page
` Toggle compact comment list view
g Toggle anti-kibitzer