Luck based medicine: inositol

Elizabeth22 Sep 2023 20:10 UTC

14 points

0 comments3 min readLW link

(acesounderglass.com)

If influence functions are not approximating leave-one-out, how are they supposed to help?

Fabien Roger22 Sep 2023 14:23 UTC

40 points

2 comments3 min readLW link

Modeling p(doom) with TrojanGDP

K. Liam Smith22 Sep 2023 14:19 UTC

−2 points

2 comments13 min readLW link

Let’s talk about Impostor syndrome in AI safety

Igor Ivanov22 Sep 2023 13:51 UTC

31 points

1 comment3 min readLW link

Fund Transit With Development

jefftk22 Sep 2023 11:10 UTC

31 points

3 comments3 min readLW link

(www.jefftk.com)

Atoms to Agents Proto-Lectures

johnswentworth22 Sep 2023 6:22 UTC

56 points

4 comments2 min readLW link

(www.youtube.com)

Would You Work Harder In The Least Convenient Possible World?

Firinn22 Sep 2023 5:17 UTC

39 points

20 comments9 min readLW link

Contra Kevin Dorst’s Rational Polarization

azsantosk22 Sep 2023 4:28 UTC

6 points

1 comment9 min readLW link

What social science research do you want to see reanalyzed?

Michael Wiebe22 Sep 2023 0:03 UTC

10 points

7 comments1 min readLW link

Immortality or death by AGI

ImmortalityOrDeathByAGI21 Sep 2023 23:59 UTC

42 points

21 comments4 min readLW link

Neel Nanda on the Mechanistic Interpretability Researcher Mindset

Michaël Trazzi21 Sep 2023 19:47 UTC

34 points

1 comment3 min readLW link

(theinsideview.ai)

Require AGI to be Explainable

PeterMcCluskey21 Sep 2023 16:11 UTC

5 points

0 comments6 min readLW link

(bayesianinvestor.com)

Update to “Dominant Assurance Contract Platform”

moyamo21 Sep 2023 16:09 UTC

26 points

1 comment1 min readLW link

Sparse Autoencoders: Future Work

Logan Riggs and Aidan Ewart

21 Sep 2023 15:30 UTC

13 points

0 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

97 points

5 comments5 min readLW link

There should be more AI safety orgs

Marius Hobbhahn21 Sep 2023 14:53 UTC

116 points

6 comments17 min readLW link

[Question] How are rationalists or orgs blocked, that you can see?

Nathan Young21 Sep 2023 2:37 UTC

7 points

2 comments1 min readLW link

Notes on ChatGPT’s “memory” for strings and for events

Bill Benzon20 Sep 2023 18:12 UTC

3 points

0 comments10 min readLW link

Belief and the Truth

Sam I am20 Sep 2023 17:38 UTC

2 points

12 comments5 min readLW link

(open.substack.com)

Image Hijacks: Adversarial Images can Control Generative Models at Runtime

Scott Emmons, Luke Bailey and Euan Ong

20 Sep 2023 15:23 UTC

55 points

8 comments1 min readLW link

(arxiv.org)