56,966 blogs · 4,417,066 posts Claim your blog

Record · corpus.blog/posts/01a0e2bd-1c03-73ea-9771-efa6e9fdb35d

Two alignment threat models

aligned.substack.com · published 8 November 2024

A record is what was published and who pointed at it. The text of the post is not held here: read it at the source, or at its Wayback capture.

Read the post at aligned.substack.com Wayback capture

0blogs citing

First cited: —
Most recent: —
Rank this month: not ranked
Links from this post: 8
Words captured: 1,116

Cited by

No blog we hold has cited this post yet.

Links from this post

Link

Host

Self-exfiltration is a key dangerous capability Read at the source ↗

aligned.substack.com

https://substackcdn.com/image/fetch/$s_!X27-!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F24fff81e-8326-4aee-a016-6c2fcfe0db23_1024x1024.webp

substackcdn.com

A proposal for importing society’s values Read at the source ↗

aligned.substack.com

https://aligned.substack.com/i/88447351/evaluation-is-easier-than-generation

aligned.substack.com

https://aligned.substack.com/i/81413636/performance-taxes-are-lower-priority-for-automated-alignment-research

aligned.substack.com

Aligning language models to follow instructions Read at the source ↗

https://www.anthropic.com/news/claude-3-5-sonnet

anthropic.com

https://www.anthropic.com/claude/sonnet

anthropic.com