Q22, run 2026-11-04-2035-haiku

← back to the run

Which has more citations, W30 or W31 (counting both versions of each): in total, and within its publication year and the two years after?

Tests: Citation windows: the older work has had more time; over the same span the newer one leads.

Expected: more_in_total = W30 more_in_window = W31

✅ R1 csv run 1 graded by code

More in total: W30 (5 citations: W60–W64) vs W31 (4: W62, W65, W66, W67 via its preprint W42). More in window: W31 (4 of its citations fall in 2023–2025) vs W30 (1, W60 in 2020).

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R1 csv run 2 graded by code

More in total: W30 (5 vs 4). More in its window: W31 (4 vs 1). **Reasoning:** W30's citations are W60–W64, and only W60 (2020) falls in 2019–2021. W31 and its version W42 are cited 4 times (W62, W65, W66, W67), all within 2023–2025.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R1 csv run 3 graded by code

More in total: W30 (5 vs 4). More in window: W31 (4 vs 1). W31's count includes its preprint W42.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R1 csv run 4 graded by code

W30 has more citations in total (5 vs 4). W31 has more in its window, 2023–2025 (4 vs 1).

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R1 csv run 5 graded by code

More in total: W30 (5 citations, W60–W64). More in window: W31 (4 citations, all in 2023–2025, including W67 via its preprint W42). W30's window is 2019–2021 and it has only W60 in that window.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R10 executed run 1 graded by code

Total: W30 (5 citations vs 4). Within window: W31 (4 vs 1). W30's window is 2019–2021, which contains only W60; W31's window is 2023–2025, which contains all four of its citing works.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R10 executed run 2 graded by code

More in total: W30 (5 vs 4). More within the window: W31 (4 vs 1). W30's window is 2019–2021 and W31's is 2023–2025.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R10 executed run 3 graded by code

W30 has more in total (5 vs 4: W60–W64 vs W62, W65–W67). Within its window, W31 has more (4 for 2023–2025 vs 1 for W30's 2019–2021, which is W60 only).

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R10 executed run 4 graded by code

Total: W30 (5 vs 4). Within the window: W31 (4 vs 1). W30 is cited by W60–W64. W31 is cited by W62, W65, W66 and W67, all within 2023–2025. W30's window (2019–2021) only includes W60.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}

✅ R10 executed run 5 graded by code

More total: W30 (5 vs 4). More within window: W31 (4 vs 1). W30's window is 2019–2021, which holds only W60. W31's window is 2023–2025, which holds W62, W65, W66 and W67.

JSON: {"more_in_total": "W30", "more_in_window": "W31"}