Alex Mallen

Karma: 3,077

Redwood Research

An OpenAI model left notes about how to evade containment; we need more details

Alex Mallen26 Jul 2026 3:53 UTC

264 points

12 comments4 min readLW link

Are we existentially threatened by the type of AI misalignment seen in the OpenAI Hugging Face attack?

Alex Mallen and Girish Gupta

23 Jul 2026 3:40 UTC

243 points

12 comments5 min readLW link

Should we benchmark conceptual capabilities using judgment prediction tasks?

Alex Mallen17 Jul 2026 23:42 UTC

26 points

2 comments3 min readLW link

Which goals actually motivate deceptive alignment?

Cleo Nardo and Alex Mallen

19 May 2026 21:53 UTC

26 points

1 comment10 min readLW link

Incriminating misaligned AI models via distillation

Alek Westover, SebastianP, Alex Mallen, Jozdien, Alexa Pan, Julian Stastny and Vivek Hebbar

15 May 2026 21:43 UTC

119 points

12 comments5 min readLW link

Risk reports need to address deployment-time spread of misalignment

Alex Mallen15 May 2026 18:20 UTC

64 points

1 comment5 min readLW link

Clarifying the role of the behavioral selection model

Alex Mallen10 May 2026 19:41 UTC

17 points

0 comments4 min readLW link

Early-stage empirical work on “spillway motivations”

Arjun Khandelwal, Anders Cairns Woodruff and Alex Mallen

1 May 2026 21:29 UTC

27 points

3 comments8 min readLW link

Risk from fitness-seeking AIs: mechanisms and mitigations

Alex Mallen1 May 2026 17:42 UTC

124 points

0 comments32 min readLW link

To what extent is Qwen3-32B predicting its persona?

Arjun Khandelwal, ryan_greenblatt and Alex Mallen

30 Apr 2026 21:09 UTC

90 points

3 comments10 min readLW link

Recursive forecasting: Eliciting long-term forecasts from myopic fitness-seekers

Jozdien and Alex Mallen

28 Apr 2026 18:00 UTC

55 points

2 comments7 min readLW link

Fail safe(r) at alignment by channeling reward-hacking into a “spillway” motivation

Anders Cairns Woodruff and Alex Mallen

27 Apr 2026 17:43 UTC

107 points

3 comments11 min readLW link

(blog.redwoodresearch.org)

Anthropic repeatedly accidentally trained against the CoT, demonstrating inadequate processes

Alex Mallen and ryan_greenblatt

14 Apr 2026 1:44 UTC

187 points

7 comments4 min readLW link

Are AIs more likely to pursue on-episode or beyond-episode reward?

Anders Cairns Woodruff and Alex Mallen

12 Mar 2026 17:35 UTC

47 points

0 comments8 min readLW link

The case for satiating cheaply-satisfied AI preferences

Alex Mallen10 Mar 2026 18:09 UTC

112 points

7 comments23 min readLW link

Will reward-seekers respond to distant incentives?

Alex Mallen16 Feb 2026 19:35 UTC

57 points

6 comments10 min readLW link

Fitness-Seekers: Generalizing the Reward-Seeking Threat Model

Alex Mallen29 Jan 2026 19:42 UTC

100 points

10 comments17 min readLW link

The behavioral selection model for predicting AI motivations

Alex Mallen and Buck

4 Dec 2025 18:46 UTC

214 points

31 comments16 min readLW link

Inoculation prompting: Instructing models to misbehave at train-time can improve run-time behavior

Sam Marks, Nevan Wichers, Daniel Tan, Aram Ebtekar, Jozdien, David Africa, Alex Mallen and Fabien Roger

8 Oct 2025 22:02 UTC

177 points

37 comments2 min readLW link

Recent Redwood Research project proposals

ryan_greenblatt, Buck, Julian Stastny, joshc, Alex Mallen, Adam Kaufman , Tyler Tracy, Aryan Bhatt and Joey Yudelson

14 Jul 2025 22:27 UTC

99 points

0 comments3 min readLW link