Political sycophancy as a model organism of scheming

Alex Mallen and Vivek Hebbar

May 12, 2025, 5:49 PM

26 points

0 comments14 min readLW link

AIs at the current capability level may be important for future safety work

ryan_greenblattMay 12, 2025, 2:06 PM

57 points

1 comment4 min readLW link

Highly Opinionated Advice on How to Write ML Papers

Neel NandaMay 12, 2025, 1:59 AM

35 points

2 comments32 min readLW link

Absolute Zero: Alpha Zero for LLM

alapmi11 May 2025 20:42 UTC

13 points

2 comments1 min readLW link

Mind the Coherence Gap: Lessons from Steering Llama with Goodfire

eitan sprejer9 May 2025 21:29 UTC

4 points

0 comments6 min readLW link

Slow corporations as an intuition pump for AI R&D automation

ryan_greenblatt and elifland

9 May 2025 14:49 UTC

88 points

21 comments9 min readLW link

Video & transcript: Challenges for Safe & Beneficial Brain-Like AGI

Steven Byrnes8 May 2025 21:11 UTC

24 points

0 comments18 min readLW link

Misalignment and Strategic Underperformance: An Analysis of Sandbagging and Exploration Hacking

Buck and Julian Stastny

8 May 2025 19:06 UTC

75 points

1 comment15 min readLW link

An alignment safety case sketch based on debate

Marie_DB, Jacob Pfau, Benjamin Hilton and Geoffrey Irving

8 May 2025 15:02 UTC

55 points

13 comments25 min readLW link

(arxiv.org)

UK AISI’s Alignment Team: Research Agenda

Benjamin Hilton, Jacob Pfau, Marie_DB and Geoffrey Irving

7 May 2025 16:33 UTC

107 points

2 comments10 min readLW link

The Sweet Lesson: AI Safety Should Scale With Compute

Jesse Hoogland5 May 2025 19:03 UTC

86 points

1 comment3 min readLW link

Interpretability Will Not Reliably Find Deceptive AI

Neel Nanda4 May 2025 16:32 UTC

257 points

30 comments7 min readLW link

SimpleStories: A Better Synthetic Dataset and Tiny Models for Interpretability

Lennart Finke3 May 2025 14:04 UTC

11 points

0 comments1 min readLW link

Interim Research Report: Mechanisms of Awareness

Josh Engels, Neel Nanda and Senthooran Rajamanoharan

2 May 2025 20:29 UTC

38 points

5 comments8 min readLW link

What’s going on with AI progress and trends? (As of 5/2025)

ryan_greenblatt2 May 2025 19:00 UTC

70 points

7 comments8 min readLW link

My Research Process: Understanding and Cultivating Research Taste

Neel Nanda1 May 2025 23:08 UTC

26 points

1 comment9 min readLW link

What is Inadequate about Bayesianism for AI Alignment: Motivating Infra-Bayesianism

Brittany Gelb1 May 2025 19:06 UTC

17 points

0 comments7 min readLW link

How can we solve diffuse threats like research sabotage with AI control?

Vivek Hebbar30 Apr 2025 19:23 UTC

43 points

0 comments8 min readLW link

Video and transcript of talk on automating alignment research

Joe Carlsmith30 Apr 2025 17:43 UTC

21 points

0 comments24 min readLW link

(joecarlsmith.com)

Can we safely automate alignment research?

Joe Carlsmith30 Apr 2025 17:37 UTC

53 points

29 comments48 min readLW link

(joecarlsmith.com)