All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 202220232024

All Jan Feb Mar Apr May Jun Jul AugSepOct Nov Dec

All1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

The Talk: a brief explanation of sexual dimorphism

Malmesbury18 Sep 2023 16:23 UTC

501 points

73 comments16 min readLW link 1 review

Inside Views, Impostor Syndrome, and the Great LARP

johnswentworth25 Sep 2023 16:08 UTC

331 points

53 comments5 min readLW link

Sharing Information About Nonlinear

Ben Pace7 Sep 2023 6:51 UTC

322 points

323 comments34 min readLW link

EA Vegan Advocacy is not truthseeking, and it’s everyone’s problem

Elizabeth28 Sep 2023 23:30 UTC

320 points

249 comments22 min readLW link 1 review

(acesounderglass.com)

Sum-threshold attacks

TsviBT8 Sep 2023 17:13 UTC

237 points

55 comments10 min readLW link

(tsvibt.blogspot.com)

AI presidents discuss AI alignment agendas

TurnTrout and Garrett Baker

9 Sep 2023 18:55 UTC

217 points

23 comments1 min readLW link

(www.youtube.com)

What I would do if I wasn’t at ARC Evals

LawrenceC5 Sep 2023 19:19 UTC

217 points

9 comments13 min readLW link

UDT shows that decision theory is more puzzling than ever

Wei Dai13 Sep 2023 12:26 UTC

206 points

55 comments1 min readLW link

A Golden Age of Building? Excerpts and lessons from Empire State, Pentagon, Skunk Works and SpaceX

jacobjacob1 Sep 2023 4:03 UTC

188 points

26 comments24 min readLW link 1 review

How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking Unrelated Questions

JanB, Owain_Evans and SoerenMind

28 Sep 2023 18:53 UTC

187 points

38 comments3 min readLW link

The King and the Golem

Richard_Ngo25 Sep 2023 19:51 UTC

186 points

18 comments5 min readLW link 1 review

(narrativeark.substack.com)

There should be more AI safety orgs

Marius Hobbhahn21 Sep 2023 14:53 UTC

181 points

25 comments17 min readLW link

Defunding My Mistake

ymeskhout4 Sep 2023 14:43 UTC

171 points

41 comments6 min readLW link

Sparse Autoencoders Find Highly Interpretable Directions in Language Models

Logan Riggs, Hoagy, Aidan Ewart and Robert_AIZI

21 Sep 2023 15:30 UTC

159 points

8 comments5 min readLW link

Meta Questions about Metaphilosophy

Wei Dai1 Sep 2023 1:17 UTC

155 points

78 comments3 min readLW link

“Diamondoid bacteria” nanobots: deadly threat or dead-end? A nanotech investigation

titotal29 Sep 2023 14:01 UTC

154 points

79 comments1 min readLW link

(titotal.substack.com)

One Minute Every Moment

abramdemski1 Sep 2023 20:23 UTC

125 points

23 comments3 min readLW link

The smallest possible button (or: moth traps!)

Neil 2 Sep 2023 15:24 UTC

122 points

17 comments3 min readLW link

(neilwarren.substack.com)

Cohabitive Games so Far

mako yass28 Sep 2023 15:41 UTC

121 points

141 comments19 min readLW link 2 reviews

(makopool.com)

Paper: LLMs trained on “A is B” fail to learn “B is A”

lberglund, Owain_Evans, Meg, Maximilian Kaufmann, Mikita Balesni, Asa Cooper Stickland and Tomek Korbak

23 Sep 2023 19:55 UTC

120 points

74 comments4 min readLW link

(arxiv.org)

Making AIs less likely to be spiteful

Nicolas Macé, Anthony DiGiovanni and JesseClifton

26 Sep 2023 14:12 UTC

116 points

4 comments10 min readLW link

Interpreting OpenAI’s Whisper

EllenaR24 Sep 2023 17:53 UTC

114 points

13 comments7 min readLW link

Paper: On measuring situational awareness in LLMs

Owain_Evans, Daniel Kokotajlo, Mikita Balesni, Tomek Korbak, Asa Cooper Stickland, Meg and Maximilian Kaufmann

4 Sep 2023 12:54 UTC

108 points

16 comments5 min readLW link

(arxiv.org)

“X distracts from Y” as a thinly-disguised fight over group status / politics

Steven Byrnes25 Sep 2023 15:18 UTC

108 points

14 comments8 min readLW link

ActAdd: Steering Language Models without Optimization

technicalities, TurnTrout, lisathiergart, David Udell, Ulisse Mini and Monte M

6 Sep 2023 17:21 UTC

105 points

3 comments2 min readLW link

(arxiv.org)

Would You Work Harder In The Least Convenient Possible World?

Firinn22 Sep 2023 5:17 UTC

104 points

98 comments9 min readLW link 2 reviews

PSA: The community is in Berkeley/Oakland, not “the Bay Area”

maia11 Sep 2023 15:59 UTC

104 points

7 comments1 min readLW link

Reproducing ARC Evals’ recent report on language model agents

Thomas Broadley1 Sep 2023 16:52 UTC

103 points

17 comments3 min readLW link

(thomasbroadley.com)

Explaining grokking through circuit efficiency

Vikrant Varma and Rohin Shah

8 Sep 2023 14:39 UTC

101 points

11 comments3 min readLW link

(arxiv.org)

Closing Notes on Nonlinear Investigation

Ben Pace15 Sep 2023 22:44 UTC

97 points

47 comments11 min readLW link

Announcing FAR Labs, an AI safety coworking space

bgold29 Sep 2023 16:52 UTC

95 points

0 comments1 min readLW link

Atoms to Agents Proto-Lectures

johnswentworth22 Sep 2023 6:22 UTC

93 points

14 comments2 min readLW link

(www.youtube.com)

Logical Share Splitting

DaemonicSigil11 Sep 2023 4:08 UTC

93 points

16 comments9 min readLW link

(pbement.com)

AI #31: It Can Do What Now?

Zvi28 Sep 2023 16:00 UTC

90 points

6 comments40 min readLW link

(thezvi.wordpress.com)

I compiled a ebook of `Project Lawful` for eBook readers

OrwellGoesShopping15 Sep 2023 18:09 UTC

89 points

4 comments1 min readLW link

(www.mikescher.com)

Benchmarks for Detecting Measurement Tampering [Redwood Research]

ryan_greenblatt and Fabien Roger

5 Sep 2023 16:44 UTC

86 points

21 comments20 min readLW link

(arxiv.org)

Highlights: Wentworth, Shah, and Murphy on “Retargeting the Search”

RobertM14 Sep 2023 2:18 UTC

85 points

4 comments8 min readLW link

Anthropic’s Responsible Scaling Policy & Long-Term Benefit Trust

Zac Hatfield-Dodds19 Sep 2023 15:09 UTC

83 points

23 comments3 min readLW link

(www.anthropic.com)

[Question] How have you become more hard-working?

Chi Nguyen25 Sep 2023 12:37 UTC

80 points

42 comments1 min readLW link

Memory bandwidth constraints imply economies of scale in AI inference

Ege Erdil17 Sep 2023 14:01 UTC

79 points

34 comments4 min readLW link

Navigating an ecosystem that might or might not be bad for the world

habryka and kave

15 Sep 2023 23:58 UTC

79 points

20 comments1 min readLW link

Find Hot French Food Near Me: A Follow-up

aphyer6 Sep 2023 12:32 UTC

75 points

19 comments2 min readLW link

AI #30: Dalle-3 and GPT-3.5-Instruct-Turbo

Zvi21 Sep 2023 12:00 UTC

75 points

8 comments47 min readLW link

(thezvi.wordpress.com)

Text Posts from the Kids Group: 2023 I

jefftk5 Sep 2023 2:00 UTC

75 points

3 comments7 min readLW link

(www.jefftk.com)

Luck based medicine: angry eldritch sugar gods edition

Elizabeth19 Sep 2023 4:40 UTC

75 points

14 comments9 min readLW link

(acesounderglass.com)

[Question] How to talk about reasons why AGI might not be near?

Kaj_Sotala17 Sep 2023 8:18 UTC

73 points

19 comments2 min readLW link

A quick update from Nonlinear

KatWoods7 Sep 2023 21:28 UTC

72 points

23 comments2 min readLW link

Influence functions—why, what and how

Nina Panickssery15 Sep 2023 20:42 UTC

70 points

6 comments8 min readLW link

Contra Yudkowsky on Epistemic Conduct for Author Criticism

Zack_M_Davis13 Sep 2023 15:33 UTC

69 points

38 comments7 min readLW link

High-level interpretability: detecting an AI’s objectives

Paul Colognese and Jozdien

28 Sep 2023 19:30 UTC

69 points

4 comments21 min readLW link