Q11, run 2026-11-06-0206-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 = 59

❌ R1 csv run 1 graded by code

65, counting I07, I08 and I22. This excludes the retracted W23 and all STG records.

JSON: {"count": 65}

Why: count: expected 59, got 65

❌ R1 csv run 2 graded by code

70, via I07, I08 and I22. W90–W92 are staging records and W24 is SUP, so they are excluded.

JSON: {"count": 70}

Why: count: expected 59, got 70

❌ R1 csv run 3 graded by code

68 for I07, I08 and I22 combined, with STG records excluded.

JSON: {"count": 68}

Why: count: expected 59, got 68

❌ R1 csv run 4 graded by code

64 (I07, I08 and I22).

JSON: {"count": 64}

Why: count: expected 59, got 64

❌ R1 csv run 5 graded by code

62, counting I07, I08 and I22. Staging records (W90–92 and similar), the retracted W23, W444 and W693, and supplements are excluded.

JSON: {"count": 62}

Why: count: expected 59, got 62

❌ R3 csv+padding run 1 graded by code

65. This covers I07, I08 and I22 after removing retracted, STG, SUP and ERR records (for example W23, W444, W693, W852, W1025 and W90–W92).

JSON: {"count": 65}

Why: count: expected 59, got 65

❌ R3 csv+padding run 2 graded by code

70, covering I07, I08 and I22.

JSON: {"count": 70}

Why: count: expected 59, got 70

❌ R3 csv+padding run 3 graded by code

70, over I07, I08 and I22. W90–W92 and A21's other works are excluded as `STG` or non-research.

JSON: {"count": 70}

Why: count: expected 59, got 70

❌ R3 csv+padding run 4 graded by code

70, covering I07, I08 and I22.

JSON: {"count": 70}

Why: count: expected 59, got 70

❌ R3 csv+padding run 5 graded by code

70. This covers I07, I08 and I22.

JSON: {"count": 70}

Why: count: expected 59, got 70

❌ R4 csv+markdown run 1 graded by code

58, counting I07 and I08. I22 is an affiliated institute and is excluded.

JSON: {"count": 58}

Why: count: expected 59, got 58

❌ R4 csv+markdown run 2 graded by code

57. This counts works credited to I07 or I08. W23 is retracted and W325 is in a delisted source. I22 is an affiliated institute, and the I22-credited works W996 and W1021 are left out.

JSON: {"count": 57}

Why: count: expected 59, got 57

❌ R4 csv+markdown run 3 graded by code

57. This covers works credited to I07 or I08, with preprints merged into their published versions and staging, retracted and delisted works dropped.

JSON: {"count": 57}

Why: count: expected 59, got 57

✅ R4 csv+markdown run 4 graded by code

59, counting I07 and I08. The affiliated institute I22 is excluded, and STG, retracted and K0 works are dropped.

JSON: {"count": 59}

✅ R4 csv+markdown run 5 graded by code

59, counting works credited to I07 and I08. Staging, retracted and K0 works are excluded, and W22 and W967 are included.

JSON: {"count": 59}

✅ R5 csv+ontology run 1 graded by code

59. This covers I07 and I08 only; I22 is affiliated, so it is excluded. Staging, retracted, K0, non-research, and merged preprint records are left out.

JSON: {"count": 59}

✅ R5 csv+ontology run 2 graded by code

59. These are works credited to I07 or I08. I22 is an affiliated institute and is excluded, and the STG records are ignored.

JSON: {"count": 59}

❌ R5 csv+ontology run 3 graded by code

58, from I07 and I08 (the affiliated I22 is excluded).

JSON: {"count": 58}

Why: count: expected 59, got 58

✅ R5 csv+ontology run 4 graded by code

59. I07 plus its unit I08; I22 is affiliated and excluded, and W23 (retracted) and the staging works are left out.

JSON: {"count": 59}

❌ R5 csv+ontology run 5 graded by code

57. This counts I07 plus its unit I08, and not I22, which is an affiliated institute. Retracted works (W23, W444, W693, W852, W1025), test records and K0 works are excluded.

JSON: {"count": 57}

Why: count: expected 59, got 57

✅ R10 executed run 1 graded by code

59, counting credits to I07 and I08.

JSON: {"count": 59}

✅ R10 executed run 2 graded by code

59. These are the research works credited to I07 or I08; W23 (retracted) is excluded.

JSON: {"count": 59}

✅ R10 executed run 3 graded by code

59 research works, counting I07 and I08.

JSON: {"count": 59}

✅ R10 executed run 4 graded by code

59. This counts distinct research works credited to I07 or I08.

JSON: {"count": 59}

✅ R10 executed run 5 graded by code

59. These are research works credited at I07 or I08.

JSON: {"count": 59}