Not directly, it seems. You can easily test this by pasting some of the more offensive tech bro speak into a fresh claude session, to have it explain what was trying to be said. The new session won't be able to help, so claude doesn't even know what claude says!
I say "not directly", because I think it probably is meaningful, if you include the adjacent hidden thinking as context. From claude's "perspective", with that context, it probably is coherent. I naively suspect this would be hard to train. During tuning, you would probably need to reward good answers interpreted without thinking context visible!
Not directly, it seems. You can easily test this by pasting some of the more offensive tech bro speak into a fresh claude session, to have it explain what was trying to be said. The new session won't be able to help, so claude doesn't even know what claude says!
I say "not directly", because I think it probably is meaningful, if you include the adjacent hidden thinking as context. From claude's "perspective", with that context, it probably is coherent. I naively suspect this would be hard to train. During tuning, you would probably need to reward good answers interpreted without thinking context visible!