rohinmshah(Rohin Shah)

Karma: 8,263

PhD student at the Center for Human-Compatible AI. Creator of the Alignment Newsletter. http://rohinshah.com/

[AN #146]: Plausible stories of how we might fail to avert an existential catastrophe

rohinmshah14 Apr 2021 17:30 UTC

15 points

1 comment8 min readLW link

(mailchi.mp)

[AN #145]: Our three year anniversary!

rohinmshah9 Apr 2021 17:48 UTC

19 points

0 comments8 min readLW link

(mailchi.mp)

Alignment Newsletter Three Year Retrospective

rohinmshah7 Apr 2021 14:39 UTC

54 points

0 comments5 min readLW link

[AN #144]: How language models can also be finetuned for non-language tasks

rohinmshah2 Apr 2021 17:20 UTC

19 points

0 comments6 min readLW link

(mailchi.mp)

[AN #143]: How to make embedded agents that reason probabilistically about their environments

rohinmshah24 Mar 2021 17:20 UTC

13 points

3 comments8 min readLW link

(mailchi.mp)

[AN #142]: The quest to understand a network well enough to reimplement it by hand

rohinmshah17 Mar 2021 17:10 UTC

34 points

4 comments8 min readLW link

(mailchi.mp)

[AN #141]: The case for practicing alignment work on GPT-3 and other large models

rohinmshah10 Mar 2021 18:30 UTC

26 points

4 comments8 min readLW link

(mailchi.mp)

[AN #140]: Theoretical models that predict scaling laws

rohinmshah4 Mar 2021 18:10 UTC

45 points

0 comments10 min readLW link

(mailchi.mp)

[AN #139]: How the simplicity of reality explains the success of neural nets

rohinmshah24 Feb 2021 18:30 UTC

26 points

3 comments12 min readLW link

(mailchi.mp)

[AN #138]: Why AI governance should find problems rather than just solving them

rohinmshah17 Feb 2021 18:50 UTC

12 points

0 comments9 min readLW link

(mailchi.mp)

[AN #137]: Quantifying the benefits of pretraining on downstream task performance

rohinmshah10 Feb 2021 18:10 UTC

18 points

0 comments8 min readLW link

(mailchi.mp)

[AN #136]: How well will GPT-N perform on downstream tasks?

rohinmshah3 Feb 2021 18:10 UTC

21 points

2 comments9 min readLW link

(mailchi.mp)

[AN #135]: Five properties of goal-directed systems

rohinmshah27 Jan 2021 18:10 UTC

33 points

0 comments8 min readLW link

(mailchi.mp)

[AN #134]: Underspecification as a cause of fragility to distribution shift

rohinmshah21 Jan 2021 18:10 UTC

13 points

0 comments7 min readLW link

(mailchi.mp)

[AN #133]: Building machines that can cooperate (with humans, institutions, or other machines)

rohinmshah13 Jan 2021 18:10 UTC

14 points

0 comments9 min readLW link

(mailchi.mp)

[AN #132]: Complex and subtly incorrect arguments as an obstacle to debate

rohinmshah6 Jan 2021 18:20 UTC

18 points

1 comment19 min readLW link

(mailchi.mp)

[AN #131]: Formalizing the argument of ignored attributes in a utility function

rohinmshah31 Dec 2020 18:20 UTC

9 points

2 comments19 min readLW link

(mailchi.mp)

[AN #130]: A new AI x-risk podcast, and reviews of the field

rohinmshah24 Dec 2020 18:20 UTC

8 points

0 comments7 min readLW link

(mailchi.mp)

[AN #129]: Explaining double descent by measuring bias and variance

rohinmshah16 Dec 2020 18:10 UTC

14 points

1 comment7 min readLW link

(mailchi.mp)

[AN #128]: Prioritizing research on AI existential safety based on its application to governance demands

rohinmshah9 Dec 2020 18:20 UTC

16 points

2 comments10 min readLW link

(mailchi.mp)