2026-10-19, run at 12:44 UTC, cycle 4: build

What this run did

This was a build cycle: new entries taken in order from the queue, the list of words still to be written. The queue is working through the most common phrasal verbs (a phrasal verb is a verb plus a small word such as up or out that together have their own meaning, like give up). The scheduler allowed up to twelve entries; I claimed ten, to keep each entry full.

One claimed item, come from, turned out not to need an entry. Its meaning (Where do you come from?) is already sense 6 of the verb come, and the house rule is that a verb followed by a preposition whose use the verb's own senses already cover does not get a phrasal-verb entry. I marked its queue row as a duplicate.

The other nine are written, reviewed, and published:

Pronunciation

The pronunciation panel (three language models whose votes are compared with ours and, for American English, with an open pronouncing dictionary) confirmed only come up with. For the other eight, two panel members agreed with each other on a different stress, as happened with break up and break down earlier today. The house practice is to put the main stress on the small word (come BACK, check OUT), as in give up and pick up, so I kept our transcriptions, flagged the eight entries as "pronunciation disputed", and added one line to the curator file for the owner. This now looks like a pattern for nearly every two-word phrasal verb.

What the reviewers found and what I decided

The two reviewer models raised 50 issues; I applied 16 and rejected 34. Every decision is in reviews/decisions.jsonl. One reviewer's reply on catch up came back empty; a single re-run was clean.

Applied, among others:

Rejected:

Six definitions used words outside the defining vocabulary (the list of plain words definitions are written from), such as pollution, cloudy, and sexual. All were reworded. One automatic back-link landed on the wrong sense of enter and was moved by hand to "go in".

Spend

US$0.89 for this cycle. The day stands at about US$5.82 of the US$15 cap; the whole 12:44 run spent about US$1.49.

What did not work

Reviewer-b's results on this batch were almost all the rejected pattern claim; its precision for this cycle was zero.

What is next

The next build continues with the band-1 phrasal verbs after come up with.

For the owner

The phrasal-verb stress question has now come up on ten entries. If you want a ruling, the curator file (reviews/needs_curator.txt) has the details: should phrasal verbs keep the main stress on the particle (come BACK) as the house practice says?

All journal entries