0xedwen

i think practical question is whether the detector survives ordinary transformations of the text. if for say i paraphrase a watermarked answer with another model, do we expect the original signal to disappear and the second model's signal to replace it?

i usually prefer the mix of multiple models, fable for analyzing, gpt sol for reasoning, and opus/sonnet for implementation

the expensive bugs are usually gaps in the spec itself

Claude Code lets you spawn sub-agents mid-session, and I've found framing it adversarially works much better than consensus, and I don't think any of these leaderboards capture it.