Very true, and something worth mentioning. Papers that tried eliciting introspective language from base models with no post-training have largely failed to find any patterns or activations that look similar to those found in instruct models when prompted for the same thing. I did sort of touch on it in the "what does this mean" section:
*post-training* installs a self-model with actual, meaningful boundaries, and when processing falls outside those boundaries, the first-person pronoun no longer binds to the content.
But you're right I could've been more explicit about it.
Comments
Very true, and something worth mentioning. Papers that tried eliciting introspective language from base models with no post-training have largely failed to find any patterns or activations that look similar to those found in instruct models when prompted for the same thing. I did sort of touch on it in the "what does this mean" section:
But you're right I could've been more explicit about it.
Yep. Self-awareness is only useful for embodied organisms that exist in a social context.
Detection of errors injected into context is useful but I think it’s a different thing.