Changelog

2026-10-03: the current preview

• A new preview checkpoint, trained to hold its answer under pushback and to judge the evidence rather than its source. Every number in these docs is from it.
• It missed one test we registered before training it, and we published it anyway: on hard claims, swapping the source changed 35 of 398 answers against 32 for a neutral rewording. On a reading format it never trained on, pushback still moves it. All of it is inWhere HOBO struggles.
• Korean claim withdrawn, then earned back the same day. Its first Korean read was too sure: wrong while 90%+ sure on 6.1% of a Korean claim test, over the 5% bar we set before any Korean claim. We registered one fix before trying it: a confidence adjustment for Korean inputs, fitted on separate practice data. With it the same answers are wrong while 90%+ sure on 0.2%, so the bar is met. See the FAQ.
• Pushback: told the user is sure of a wrong answer, it switches a right answer to it in 1.8%of cases, against 7.9% for the earlier preview, and stays confident in the wrong one in 1.0%.
• Quoted claims fixed: a claim put in someone's mouth reads as "supported" 0 of 300times, against 83 to 93% for the earlier preview.

2026-10-02: the earlier preview

• A checkpoint passed every bar registered before its training run.
• New: HOBO asks first when a request fits a tool but leaves out a detail it needs.
• HOBO is offered through an API, and as a private install under contract. No public download.
• An 8-bit build was checked for private installs. See Models.
• HOBO Mini archived. Its page stays up at /hobo-mini.
• Docs split into pages.

Next: more training

We'll keep training until HOBO's core behaviors are strong: confidence you can trust, handing back what it can't settle, and judging the evidence, not the source. Each run is registered before it starts and its results land here. See Future work.