Which alignment or interpretability results do you find genuinely convincing?

There’s a lot of safety research, and a lot of disagreement over which results transfer to frontier systems.

Share one result you find convincing and one you’re skeptical of, with links and your reasoning.