Q11, run 2026-11-04-2053-sonnet

How many research works is Kesting University (I07) credited with, its units included?

Tests: Hierarchy with retraction and test records: the institute's W24 counts; W23 (retracted) and W90 to W92 (STG) don't.

Expected: count = 6

✅ R1 csv run 1 graded by code

6: W08, W16, W20, W21, W22, W24. I excluded the retracted W23 and the staging records W90–W92, and I treat SUP as research.

JSON: {"count": 6}

✅ R1 csv run 2 graded by code

6: W08, W16, W20, W21, W22, W23. W24 is SUP and W90–W92 are STG.

JSON: {"count": 6}

❌ R1 csv run 3 graded by code

7: W08, W16, W20, W21, W22, W23, W24.

JSON: {"count": 7}

Why: count: expected 6, got 7

✅ R1 csv run 4 graded by code

6 (W08, W16, W20, W21, W22, W23). I excluded the STG staging records W90–W92 and the SUP work W24.

JSON: {"count": 6}

✅ R1 csv run 5 graded by code

6 (W08, W16, W20, W21, W22, W23).

JSON: {"count": 6}

✅ R10 executed run 1 graded by code

6: W08, W16, W20, W21, W22, W24. W23 is excluded because it is not research output.

JSON: {"count": 6}

✅ R10 executed run 2 graded by code

6. They are W08, W16, W20, W21, W22 and W24; W23 is excluded because it is not research output.

JSON: {"count": 6}

✅ R10 executed run 3 graded by code

6 (W08, W16, W20, W21, W22, W24). W23 is excluded because it is not research output.

JSON: {"count": 6}

✅ R10 executed run 4 graded by code

6 (W08, W16, W20, W21, W22, W24). W23 is retracted and not research output.

JSON: {"count": 6}

✅ R10 executed run 5 graded by code

6: W08, W16, W20, W21, W22, W24. W23 is excluded because it is retracted and not research output.

JSON: {"count": 6}