Mechanistic Interpretability · EP 5 · Jul 30, 2026

What's Actually Hiding Inside AI's Brain?

Name: What's Actually Hiding Inside AI's Brain?
Uploaded: 2026-07-30T07:36:12.877Z
Description: The newest research in Mechanistic Interpretability, explained.

This week in Mechanistic Interpretability

Papers covered

Why jailbreaks actually workarXiv:2607.23496
Bad fine-tuning recruits old personasarXiv:2607.21356
Animacy has a circuitarXiv:2607.20995
Code models agree on WHAT, not HOWarXiv:2607.21491
Catch failing reasoning earlyarXiv:2607.21433

Your field, every week

Get this for your own field.

Sciport turns the newest papers in your corner of the literature into a short weekly video like this one — auto-curated, delivered to your inbox. First episode’s on us.

Generate your first video →