Mark Xu

Karma: 4,327

I do alignment research at the Alignment Research Center. Learn more about me at markxu.com/about

Estimating Tail Risk in Neural Networks

Mark Xu13 Sep 2024 20:00 UTC

68 points

9 comments23 min readLW link

(www.alignment.org)

Backdoors as an analogy for deceptive alignment

Jacob_Hilton and Mark Xu

6 Sep 2024 15:30 UTC

104 points

2 comments8 min readLW link

(www.alignment.org)

If you weren’t such an idiot...

kave and Mark Xu

2 Mar 2024 0:01 UTC

147 points

74 comments2 min readLW link

(markxu.com)

ARC is hiring theoretical researchers

paulfchristiano, Jacob_Hilton and Mark Xu

12 Jun 2023 18:50 UTC

126 points

12 comments4 min readLW link

(www.alignment.org)

How to do theoretical research, a personal perspective

Mark Xu19 Aug 2022 19:41 UTC

91 points

6 comments15 min readLW link

ELK prize results

paulfchristiano and Mark Xu

9 Mar 2022 0:01 UTC

138 points

50 comments21 min readLW link

ELK First Round Contest Winners

Mark Xu and paulfchristiano

26 Jan 2022 2:56 UTC

65 points

6 comments1 min readLW link

ARC’s first technical report: Eliciting Latent Knowledge

paulfchristiano, Mark Xu and Ajeya Cotra

14 Dec 2021 20:09 UTC

225 points

90 comments1 min readLW link 3 reviews

(docs.google.com)

ARC is hiring!

paulfchristiano and Mark Xu

14 Dec 2021 20:09 UTC

64 points

2 comments1 min readLW link

Your Time Might Be More Valuable Than You Think

Mark Xu18 Oct 2021 0:55 UTC

56 points

10 comments6 min readLW link

(markxu.com)

The Simulation Hypothesis Undercuts the SIA/Great Filter Doomsday Argument

Mark Xu and CarlShulman

1 Oct 2021 22:23 UTC

43 points

11 comments7 min readLW link

Fractional progress estimates for AI timelines and implied resource requirements

Mark Xu and CarlShulman

15 Jul 2021 18:43 UTC

55 points

6 comments7 min readLW link

Intermittent Distillations #4: Semiconductors, Economics, Intelligence, and Technological Progress.

Mark Xu8 Jul 2021 22:14 UTC

81 points

9 comments10 min readLW link

Anthropic Effects in Estimating Evolution Difficulty

Mark Xu5 Jul 2021 4:02 UTC

12 points

2 comments3 min readLW link

An Intuitive Guide to Garrabrant Induction

Mark Xu3 Jun 2021 22:21 UTC

145 points

20 comments24 min readLW link

Rogue AGI Embodies Valuable Intellectual Property

Mark Xu and CarlShulman

3 Jun 2021 20:37 UTC

71 points

9 comments3 min readLW link

Intermittent Distillations #3

Mark Xu15 May 2021 7:13 UTC

21 points

1 comment11 min readLW link

Pre-Training + Fine-Tuning Favors Deception

Mark Xu8 May 2021 18:36 UTC

27 points

3 comments3 min readLW link

Less Realistic Tales of Doom

Mark Xu6 May 2021 23:01 UTC

113 points

13 comments4 min readLW link

Agents Over Cartesian World Models

Mark Xu and evhub

27 Apr 2021 2:06 UTC

66 points

4 comments27 min readLW link