#hallucination

We gave 44 page-head tasks to Sonnet and Haiku. Valid syntax hid a made-up currency, a self-awarded rating and an unsupported language code.

Sonnet and Haiku each got 19 of 24 social posting API questions right alone and missed the same five. A skill fixed them, though its first wording br…

Confident, plausible, wrong. The detection layers that stop invented facts from reaching users.

Sonnet wrote 16 of 23 job posts right alone, Haiku 15. With a skill, 21 and 20, though two Sonnet posts that passed alone then failed.

We gave 24 small tables to Sonnet and Haiku and told both to use only printed figures. Sonnet still computed its own on eight tables, Haiku on none.

Claude Sonnet described 24 diffs and kept to the code on almost all of them. Its few slips were a security fix and a speed-up that no changed line sh…

Claude Sonnet turned 23 sets of commits into release notes. Asked plainly, it listed internal work and invented a link host; with written rules it ne…

We had Claude cut texts to a limit and turn spec sheets into shop copy. Bare, it overran limits and added praise; with rules, it lost facts or broke…

We asked Claude Sonnet and Haiku to humanize one text in five languages. Without rules, all ten answers lost the word usually, and some invented fact…

Six texts in five languages, two Claude models. Without rules Sonnet rewrote a correct text, and Haiku mostly sent a list of fixes with no corrected…

We asked Claude Sonnet and Haiku for six cold emails with no rules. Most ran far too long, and the one draft that passed invented a statistic and a d…

Claude Sonnet and Haiku summarized six short texts in five languages from a plain request. Without rules they ran long, switched language and bent co…