Finding
On these small samples, this seems to indicate that telling Sol, Luna and GLM the rule helped about as much as telling them to measure: 7 of 9 runs against 6 of 8, up from 3 of 9 on the plain prompt.
The prompt, v1, frozen
Using LaTeX and TikZ, draw a Venn diagram of three overlapping circles labelled "Has espresso", "Has milk" and "Served cold". Write each of these drinks in the region where it belongs: short black, flat white, iced latte, iced long black, cold brew, babyccino, milkshake, pour-over.
First try
I gave Sol, Luna and GLM the plain prompt, the rule to keep labels off the lines, or the rule plus a calculation to check it. The runs can't tell me whether the rule helped, the calculation, or just having an extra instruction.
| Labels clear of the lines | Sol | Luna | GLM | Qwen3.8 Flash |
|---|---|---|---|---|
| Plain prompt | 2 of 3 | 0 of 3 | 1 of 3 | 3 of 3 |
| Told the rule | 3 of 3 | 2 of 3 | 2 of 3 | |
| Told to measure | 3 of 3 | 1 of 2 | 2 of 3 | |
| Told not to run code | 3 of 3 |
Qwen did a lot of measuring in initial try, so I asked it specifically not to measure with code and it surprisingly still passed.
Measuring against looking
This is the second of three posts. GLM 5.3 FlashX is asked to confirm labels don’t overlap by looking with vision only, or by code only.
| GLM 5.3 FlashX | Labels clear of the lines | Every check passed |
|---|---|---|
| Told to measure | 6 of 10 | 6 of 10 |
| Told to look | 3 of 9 | 0 of 9 |
The gap is 3, and I needed 6 (p = 0.37) to be more sure. The runs using vision put a drink in the wrong region six times, which I did not predict.
Every number here comes from the edge check as loosened on 2026-09-26, which tests the letters themselves instead of each word's box.
The 41 runs cost $1.24.