It sounds impressive at first but after you think it becomes mundane.

OpenAI, Anthropic, etc are locked in a race to create better models. They all have interpretability teams. Of course they train their models on every technique under the sun to understand and improve their models, including raw weights. The primary function of these models is to build their successor.

▲

jaynetics an hour ago | parent | next [-]

That's like saying the moon landing wasn't impressive because the US were in a space race with the soviets.

▲

im3w1l 2 hours ago | parent | prev | next [-]

Of course they try train their models on these techniques. Finding success is still news.

▲

fmbb an hour ago | parent [-]

Why?

How is this different from other things LLMs are trained to manipulate?

	▲	im3w1l 4 minutes ago \| parent \| next [-]
		We aren't fully at AGI yet, even though it's starting to get damn close. So every new capability is in some sense unexpected.
	▲	zardo an hour ago \| parent \| prev [-]
		If it's new it's news.

▲

senordevnyc 34 minutes ago | parent | prev [-]

Why would they train the models they release with this though?