An LLM invented a feature by hijacking my tool schema(ratnotes.substack.com)

ratnotes.substack.com

An LLM invented a feature by hijacking my tool schema

https://ratnotes.substack.com/p/i-thought-i-had-a-bug

1 comments

Post author here. Happy to answer questions and discuss further. The essay has an appendix with the model's own self-report on its reasoning (the most load-bearing evidence, IMO), so worth scrolling to the end if you're skeptical of the rest.

Curious what you'd propose as alternative explanations, especially from folks with pointers to related literature.