Jesse Hoogland

Karma: 1,918

Executive director at Timaeus. Working on singular learning theory and developmental interpretability.

Website: jessehoogland.com

Twitter: @jesse_hoogland

o1: A Technical Primer

Jesse Hoogland9 Dec 2024 19:09 UTC

146 points

17 comments9 min readLW link

(www.youtube.com)

New o1-like model (QwQ) beats Claude 3.5 Sonnet with only 32B parameters

Jesse Hoogland27 Nov 2024 22:06 UTC

68 points

4 comments1 min readLW link

(qwenlm.github.io)

Timaeus is hiring!

Jesse Hoogland, Stan van Wingerden, Alexander Gietelink Oldenziel and Daniel Murfet

12 Jul 2024 23:42 UTC

67 points

6 comments2 min readLW link

Stagewise Development in Neural Networks

Jesse Hoogland, Liam Carroll and Daniel Murfet

20 Mar 2024 19:54 UTC

90 points

1 comment11 min readLW link

Timaeus’s First Four Months

Jesse Hoogland, Daniel Murfet, Stan van Wingerden and Alexander Gietelink Oldenziel

28 Feb 2024 17:01 UTC

172 points

6 comments6 min readLW link

Generalization, from thermodynamics to statistical physics

Jesse Hoogland30 Nov 2023 21:28 UTC

63 points

9 comments28 min readLW link

Announcing Timaeus

Jesse Hoogland, Daniel Murfet, Alexander Gietelink Oldenziel and Stan van Wingerden

22 Oct 2023 11:59 UTC

187 points

15 comments4 min readLW link

You’re Measuring Model Complexity Wrong

Jesse Hoogland and Stan van Wingerden

11 Oct 2023 11:46 UTC

89 points

15 comments13 min readLW link

Open Call for Research Assistants in Developmental Interpretability

Jesse Hoogland, Daniel Murfet, Alexander Gietelink Oldenziel and Stan van Wingerden

30 Aug 2023 9:02 UTC

55 points

11 comments4 min readLW link

Apply for the 2023 Developmental Interpretability Conference!

Stan van Wingerden, Alexander Gietelink Oldenziel, Jesse Hoogland and Daniel Murfet

25 Aug 2023 7:12 UTC

33 points

0 comments2 min readLW link

Towards Developmental Interpretability

Jesse Hoogland, Alexander Gietelink Oldenziel, Daniel Murfet and Stan van Wingerden

12 Jul 2023 19:33 UTC

180 points

10 comments9 min readLW link 1 review

Approximation is expensive, but the lunch is cheap

Jesse Hoogland and Zach Furman

19 Apr 2023 14:19 UTC

70 points

3 comments16 min readLW link

Jesse Hoogland’s Shortform

Jesse Hoogland12 Apr 2023 9:32 UTC

4 points

15 comments1 min readLW link

Singularities against the Singularity: Announcing Workshop on Singular Learning Theory and Alignment

Jesse Hoogland, Alexander Gietelink Oldenziel and Daniel Murfet

1 Apr 2023 9:58 UTC

87 points

0 comments1 min readLW link

(singularlearningtheory.com)

Empirical risk minimization is fundamentally confused

Jesse Hoogland22 Mar 2023 16:58 UTC

32 points

5 comments1 min readLW link

The shallow reality of ‘deep learning theory’

Jesse Hoogland22 Feb 2023 4:16 UTC

34 points

11 comments3 min readLW link

(www.jessehoogland.com)

Gradient surfing: the hidden role of regularization

Jesse Hoogland6 Feb 2023 3:50 UTC

37 points

9 comments14 min readLW link

(www.jessehoogland.com)

Spooky action at a distance in the loss landscape

Jesse Hoogland and Filip Sondej

28 Jan 2023 0:22 UTC

61 points

4 comments7 min readLW link

(www.jessehoogland.com)

Neural networks generalize because of this one weird trick

Jesse Hoogland18 Jan 2023 0:10 UTC

179 points

29 comments53 min readLW link 1 review

(www.jessehoogland.com)

No, human brains are not (much) more efficient than computers

Jesse Hoogland6 Sep 2022 13:53 UTC

22 points

21 comments3 min readLW link

(www.jessehoogland.com)