luciaquirke comments on It’s Owl in the Numbers: Token Entanglement in Subliminal Learning

[luciaquirke]

I used your prompt experiment to get a similar result to the emergent misalignment finetuning paper. Unedited replication log: https://​​docs.google.com/​​document/​​d/​​1-oZ4PnxpVca_AZ0UY5tQK2GANO89UIFyTz5oBzEXnMg/​​edit?usp=sharing