Sin­gu­lar Learn­ing The­ory for Dummies

Rahul ChandOct 15, 2024, 9:13 PM
1 point
0 comments8 min readLW link

Distil­la­tion Of Deep­Seek-Prover V1.5

IvanLinOct 15, 2024, 6:53 PM
4 points
1 comment3 min readLW link

Im­prov­ing Model-Writ­ten Evals for AI Safety Benchmarking

Oct 15, 2024, 6:25 PM
30 points
0 comments18 min readLW link

Tak­ing non­log­i­cal con­cepts seriously

Kris BrownOct 15, 2024, 6:16 PM
7 points
5 comments18 min readLW link
(topos.site)

Rashomon—A news­bet­ting site

ideastheteOct 15, 2024, 6:15 PM
23 points
8 comments1 min readLW link

On the Prac­ti­cal Ap­pli­ca­tions of Interpretability

Nick JiangOct 15, 2024, 5:18 PM
4 points
1 comment7 min readLW link

An­thropic’s up­dated Re­spon­si­ble Scal­ing Policy

Zac Hatfield-DoddsOct 15, 2024, 4:46 PM
38 points
3 comments3 min readLW link
(www.anthropic.com)

[Question] When is re­ward ever the op­ti­miza­tion tar­get?

Noosphere89Oct 15, 2024, 3:09 PM
37 points
17 comments1 min readLW link

An Opinionated Evals Read­ing List

Oct 15, 2024, 2:38 PM
65 points
0 comments13 min readLW link
(www.apolloresearch.ai)

An­thropic rewrote its RSP

Zach Stein-PerlmanOct 15, 2024, 2:25 PM
46 points
19 comments6 min readLW link

[In­tu­itive self-mod­els] 5. Dis­so­ci­a­tive Iden­tity (Mul­ti­ple Per­son­al­ity) Disorder

Steven ByrnesOct 15, 2024, 1:31 PM
59 points
7 comments11 min readLW link

Eco­nomics Roundup #4

ZviOct 15, 2024, 1:20 PM
19 points
4 comments25 min readLW link
(thezvi.wordpress.com)

[Question] Is School of Thought re­lated to the Ra­tion­al­ity Com­mu­nity?

Shoshannah TekofskyOct 15, 2024, 12:41 PM
7 points
12 comments1 min readLW link

In­verse Prob­lems In Every­day Life

silentbobOct 15, 2024, 11:42 AM
14 points
2 comments8 min readLW link

Think­ing LLMs: Gen­eral In­struc­tion Fol­low­ing with Thought Generation

Bogdan Ionut CirsteaOct 15, 2024, 9:21 AM
7 points
0 comments1 min readLW link
(arxiv.org)

Thoughts On the Na­ture of Ca­pa­bil­ity Elic­i­ta­tion via Fine-tuning

Theodore ChapmanOct 15, 2024, 8:39 AM
8 points
0 comments8 min readLW link

Min­i­mal Mo­ti­va­tion of Nat­u­ral Latents

Oct 14, 2024, 10:51 PM
46 points
14 comments3 min readLW link

How long should poli­ti­cal (and other) terms be?

ohmurphyOct 14, 2024, 9:38 PM
5 points
0 comments1 min readLW link
(ohmurphy.substack.com)

Ex­am­ples of How I Use LLMs

jefftkOct 14, 2024, 5:10 PM
31 points
2 comments2 min readLW link
(www.jefftk.com)

It’s im­por­tant to know when to stop: Mechanis­tic Ex­plo­ra­tion of Gemma 2 List Generation

Gerard BoxoOct 14, 2024, 5:04 PM
9 points
0 comments6 min readLW link
(gboxo.github.io)

[Question] LW re­sources on child­hood ex­pe­riences?

nahir91595Oct 14, 2024, 5:04 PM
10 points
7 comments1 min readLW link

Free Will, Neu­rotyp­i­cal Dom­i­nance, and the Path to ASI and Neu­ral­inks: Evolv­ing Beyond Scarcity

j_passeriOct 14, 2024, 4:54 PM
−2 points
3 comments3 min readLW link

Break­throughs, Neu­ro­di­ver­gence, and Work­ing Out­side the System

j_passeriOct 14, 2024, 4:54 PM
1 point
3 comments2 min readLW link

The case for un­learn­ing that re­moves in­for­ma­tion from LLM weights

Fabien RogerOct 14, 2024, 2:08 PM
96 points
18 comments6 min readLW link

Cir­cuits in Su­per­po­si­tion: Com­press­ing many small neu­ral net­works into one

Oct 14, 2024, 1:06 PM
130 points
9 comments13 min readLW link

Beyond Defen­sive Technology

ejk64Oct 14, 2024, 11:34 AM
11 points
1 comment10 min readLW link

Why Stop AI is bar­ri­cad­ing OpenAI

RemmeltOct 14, 2024, 7:12 AM
−16 points
32 commentsLW link
(docs.google.com)

The Ex­plore vs. Ex­ploit Dilemma

nathanjzhaoOct 14, 2024, 6:20 AM
1 point
0 comments1 min readLW link
(nathanzhao.cc)

AI Align­ment via Slow Sub­strates: Early Em­piri­cal Re­sults With StarCraft II

Lester LeongOct 14, 2024, 4:05 AM
60 points
9 comments12 min readLW link

some ques­tion­able space launch guns

bhauthOct 13, 2024, 10:52 PM
17 points
0 comments4 min readLW link
(bhauth.com)

[Question] What are your fa­vorite books or blogs that are out of print, or whose do­mains have ex­pired (es­pe­cially if they also aren’t on LibGen/​Way­back/​etc, or on Ama­zon)?

Arjun PanicksseryOct 13, 2024, 8:21 PM
13 points
4 comments1 min readLW link

The Hopium Wars: the AGI En­tente Delusion

Max TegmarkOct 13, 2024, 5:00 PM
226 points
60 comments9 min readLW link

Parental Writ­ing Selec­tion Bias

jefftkOct 13, 2024, 2:00 PM
52 points
3 comments1 min readLW link
(www.jefftk.com)

Per­sonal Philosophy

XorOct 13, 2024, 3:01 AM
3 points
0 comments2 min readLW link

Con­ta­gious Beliefs—Si­mu­lat­ing Poli­ti­cal Alignment

James Stephen BrownOct 13, 2024, 12:27 AM
8 points
0 comments2 min readLW link
(nonzerosum.games)

Bi­nary en­cod­ing as a sim­ple ex­plicit con­struc­tion for superposition

tailcalledOct 12, 2024, 9:18 PM
12 points
0 comments1 min readLW link

[Question] How Should We Use Limited Time to Max­i­mize Long-Term Im­pact?

queeliusOct 12, 2024, 8:02 PM
10 points
3 comments1 min readLW link

A Per­centage Model of a Person

SableOct 12, 2024, 5:55 PM
37 points
5 comments9 min readLW link
(affablyevil.substack.com)

AI Com­pute gov­er­nance: Ver­ify­ing AI chip location

FarhanOct 12, 2024, 5:36 PM
6 points
0 comments6 min readLW link

Ge­offrey Hin­ton on the Past, Pre­sent, and Fu­ture of AI

Stephen McAleeseOct 12, 2024, 4:41 PM
22 points
5 comments18 min readLW link

[Question] I = W/​T?

HNXOct 12, 2024, 3:15 PM
0 points
3 comments1 min readLW link

AI re­search as­sis­tants com­pe­ti­tion 2024Q3: Tie be­tween Elicit and You.com

ElizabethOct 12, 2024, 3:10 PM
64 points
4 comments3 min readLW link
(acesounderglass.com)

SAE fea­tures for re­fusal and syco­phancy steer­ing vectors

Oct 12, 2024, 2:54 PM
29 points
4 comments7 min readLW link

Prices are Bounties

Maxwell TabarrokOct 12, 2024, 2:51 PM
51 points
13 comments2 min readLW link
(www.maximum-progress.com)

Differ­en­tial knowl­edge interconnection

Roman LeventovOct 12, 2024, 12:52 PM
6 points
0 comments7 min readLW link

Most ar­gu­ments for AI Doom are ei­ther bad or weak

Logan ZoellnerOct 12, 2024, 11:57 AM
4 points
101 comments3 min readLW link

Kas­sel ACX/​LW Meetup

Fernand0Oct 12, 2024, 7:47 AM
2 points
0 comments1 min readLW link

Neu­ral Net­work And New­ton’s Se­cond Law

Max MaOct 12, 2024, 6:25 AM
−10 points
0 comments1 min readLW link

[Question] If the DoJ goes through with the Google breakup,where does Deep­mind end up?

O OOct 12, 2024, 5:06 AM
5 points
1 comment1 min readLW link

My mo­ti­va­tion and the­ory of change for work­ing in AI healthtech

Andrew_CritchOct 12, 2024, 12:36 AM
178 points
39 comments14 min readLW link