As an employee of the European AI Office, it's important for me to emphasize this point: The views and opinions of the author expressed herein are personal and do not necessarily reflect those of the European Commission or other EU institutions.

Also, to stave off a common confusion: I worked at ARC Theory, which is now simply called ARC, on Paul Christiano's theoretical alignment agenda. The more famous ARC Evals was a different group working on evaluations, their work was completely separate from ARC Theory, and they were only housed under the same organization out of convenience, until ARC Evals spun off under the name METR. Nothing I write here has any implication about the work of ARC Evals/METR in any way.

Mechanistic Anomaly Detection

This is my second post in a sequence of posts on ARC's agenda. You should read the first post before this one for context.

[...]

---

Outline:

(00:55) Mechanistic Anomaly Detection

(02:54) Special case: Safe Distillation

(07:29) General case: Handling out of distribution events

(10:05) MAD solution proposal: The fragility of sensor tampering

(16:02) Detecting high-stakes failures

The original text contained 10 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.