Datasets:
dataset card
Browse files
README.md
CHANGED
|
@@ -98,6 +98,11 @@ issues in v1.0").
|
|
| 98 |
alone scores MRR@10 0.989) and a 0–180-day date window lifts BM25 from 0.547 to 0.760. Dates may be used only for the
|
| 99 |
as-of filters of the citation tasks.
|
| 100 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 101 |
Splits are temporal (train 2014–2021, dev 2022–2023, test 2024–2026); 2015 questions are dated from their XML
|
| 102 |
header.
|
| 103 |
|
|
|
|
| 98 |
alone scores MRR@10 0.989) and a 0–180-day date window lifts BM25 from 0.547 to 0.760. Dates may be used only for the
|
| 99 |
as-of filters of the citation tasks.
|
| 100 |
|
| 101 |
+
**Baseline protocol caveat.** Fine-tuning on the training split with BM25 hard negatives mined across the whole
|
| 102 |
+
archive (the pre-registered protocol) hurts (MRR@10 0.433 vs 0.490 zero-shot), consistent with a temporal
|
| 103 |
+
negative-selection shortcut; date-matched negatives give 0.632, a post-hoc single-seed result to be confirmed on the
|
| 104 |
+
next test period. Best pre-registered baseline: RRF(BM25, BGE-M3), 0.588.
|
| 105 |
+
|
| 106 |
Splits are temporal (train 2014–2021, dev 2022–2023, test 2024–2026); 2015 questions are dated from their XML
|
| 107 |
header.
|
| 108 |
|