Aug 12, 2024, 10:23 PM

42 points

14 comments21 min readLW link

Extracting SAE task features for in-context learning

Dmitrii Kharlapenko, neverix, Neel Nanda and Arthur Conmy

Aug 12, 2024, 8:34 PM

31 points

1 comment9 min readLW link

Hyppotherapy

Marius Adrian NicoarăAug 12, 2024, 8:07 PM

−3 points

0 comments1 min readLW link

Californians, tell your reps to vote yes on SB 1047!

Holly_ElmoreAug 12, 2024, 7:50 PM

40 points

24 comments LW link

[LDSL#5] Comparison and magnitude/diminishment

tailcalledAug 12, 2024, 6:47 PM

24 points

0 comments2 min readLW link

In Defense of Open-Minded UDT

abramdemskiAug 12, 2024, 6:27 PM

79 points

28 comments11 min readLW link

Humanity isn’t remotely longtermist, so arguments for AGI x-risk should focus on the near term

Seth HerdAug 12, 2024, 6:10 PM

46 points

10 comments1 min readLW link

Creating a “Conscience Calculator” to Guard-Rail an AGI

sweenesmAug 12, 2024, 4:03 PM

−2 points

0 comments13 min readLW link

Shifting Headspaces—Transitional Beast-Mode

Jonathan MoregårdAug 12, 2024, 1:02 PM

37 points

9 comments2 min readLW link

(honestliving.substack.com)

Simultaneous Footbass and Footdrums II

jefftkAug 11, 2024, 11:50 PM

9 points

0 comments1 min readLW link

(www.jefftk.com)

CultFrisbee

GauraventhAug 11, 2024, 9:36 PM

16 points

3 comments1 min readLW link

(y1d2.com)

Pleasure and suffering are not conceptual opposites

MichaelStJulesAug 11, 2024, 6:32 PM

7 points

0 comments LW link

Computational irreducibility challenges the simulation hypothesis

Clément LAug 11, 2024, 4:14 PM

4 points

17 comments7 min readLW link

[LDSL#4] Root cause analysis versus effect size estimation

tailcalledAug 11, 2024, 4:12 PM

29 points

0 comments2 min readLW link

Closed to Interpretation

Yeshua GodAug 11, 2024, 3:51 PM

−18 points

0 comments2 min readLW link

 Theories of Knowledge

Zero ContradictionsAug 11, 2024, 8:55 AM

−1 points

5 comments1 min readLW link

(thewaywardaxolotl.blogspot.com)

Unnatural abstractions

AprillionAug 10, 2024, 10:31 PM

3 points

3 comments4 min readLW link

(peter.hozak.info)

[LDSL#3] Information-orientation is in tension with magnitude-orientation

tailcalledAug 10, 2024, 9:58 PM

33 points

2 comments3 min readLW link

The AI regulator’s toolbox: A list of concrete AI governance practices

Adam JonesAug 10, 2024, 9:15 PM

9 points

1 comment34 min readLW link

(adamjones.me)

Diffusion Guided NLP: better steering, mostly a good thing

Nathan Helm-BurgerAug 10, 2024, 7:49 PM

13 points

0 comments1 min readLW link

(arxiv.org)

Tall tales and long odds

Solenoid_EntityAug 10, 2024, 3:22 PM

11 points

0 comments5 min readLW link

The Great Organism Theory of Evolution

rogersbaconAug 10, 2024, 12:26 PM

20 points

0 comments6 min readLW link

(www.secretorum.life)

Emergence, The Blind Spot of GenAI Interpretability?

Quentin FEUILLADE--MONTIXIAug 10, 2024, 10:07 AM

16 points

8 comments3 min readLW link

Rowing vs steering

Saul MunnAug 10, 2024, 7:00 AM

43 points

2 comments6 min readLW link

(www.brasstacks.blog)

Overpopulation FAQs

Zero ContradictionsAug 10, 2024, 4:21 AM

−12 points

7 comments1 min readLW link

(zerocontradictions.net)

Fermi Estimating How Long an Algorithm Takes

SatvikBeriAug 10, 2024, 1:34 AM

1 point

0 comments2 min readLW link

What’s so special about likelihoods?

mfattAug 10, 2024, 1:07 AM

6 points

1 comment1 min readLW link

Provably Safe AI: Worldview and Projects

Ben Goldhaber and Steve_Omohundro

Aug 9, 2024, 11:21 PM

54 points

44 comments7 min readLW link

All The Latest Human tFUS Studies

sarahconstantinAug 9, 2024, 10:20 PM

46 points

2 comments8 min readLW link

(sarahconstantin.substack.com)

But Where do the Variables of my Causal Model come from?

DalcyAug 9, 2024, 10:07 PM

38 points

1 comment8 min readLW link

[LDSL#2] Latent variable models, network models, and linear diffusion of sparse lognormals

tailcalledAug 9, 2024, 7:57 PM

26 points

2 comments3 min readLW link

“Which Future Mind is Me?” Is a Question of Values

dadadarrenAug 9, 2024, 6:17 PM

17 points

12 comments3 min readLW link

Simulation-aware causal decision theory: A case for one-boxing in CDT

kongus_bongusAug 9, 2024, 6:09 PM

15 points

11 comments2 min readLW link

The gatekeepers of the AI algorithms will shape our society

ProfessorFalkenAug 9, 2024, 5:59 PM

−5 points

3 comments1 min readLW link

[Question] Practical advice for secure virtual communication post easy AI voice-cloning?

hmysAug 9, 2024, 5:32 PM

6 points

5 comments2 min readLW link

The Carnot Engine of Economics

StrivingForLegibilityAug 9, 2024, 3:59 PM

5 points

0 comments5 min readLW link

Outrage Bonding

Jonathan MoregårdAug 9, 2024, 1:46 PM

63 points

12 comments2 min readLW link

(honestliving.substack.com)

Non-Transactional Compliments

Jonathan MoregårdAug 9, 2024, 1:42 PM

16 points

0 comments1 min readLW link

(honestliving.substack.com)

FarmKind’s Illusory Offer

jefftkAug 9, 2024, 11:30 AM

71 points

5 comments3 min readLW link

(www.jefftk.com)

Altruism and Selfishness

Zero ContradictionsAug 9, 2024, 6:15 AM

−3 points

0 comments1 min readLW link

(thewaywardaxolotl.blogspot.com)

Michael Streamlines on Buddhism

Chris_LeongAug 9, 2024, 4:44 AM

8 points

0 comments1 min readLW link

(x.com)

Altruism and Vitalism Aren’t Fellow Travelers

Arjun PanicksseryAug 9, 2024, 2:01 AM

24 points

2 comments3 min readLW link

(arjunpanickssery.substack.com)

Caring only about your child may increase human suffering

CipollaAug 8, 2024, 11:47 PM

−15 points

2 comments8 min readLW link

The Hessian rank bounds the learning coefficient

Lucius BushnaqAug 8, 2024, 8:55 PM

68 points

10 comments4 min readLW link

GPT-4o System Card

Zach Stein-PerlmanAug 8, 2024, 8:30 PM

68 points

11 comments2 min readLW link

(openai.com)

Parasites (not a metaphor)

lemonhopeAug 8, 2024, 8:07 PM

133 points

19 comments1 min readLW link

Some Unorthodox Ways To Achieve High GDP Growth

johnswentworth and David Lorell

Aug 8, 2024, 6:58 PM

57 points

6 comments6 min readLW link

You can remove GPT2’s LayerNorm by fine-tuning for an hour

StefanHexAug 8, 2024, 6:33 PM

165 points

11 comments8 min readLW link

Leaving MIRI, Seeking Funding

abramdemskiAug 8, 2024, 6:32 PM

264 points

19 comments2 min readLW link

[Question] Does VETLM solve AI superalignment?

Oleg TrottAug 8, 2024, 6:22 PM

−1 points

10 comments1 min readLW link