dac-research/longbench_synthetic_v4_1

LongBench Synthetic V4.1 Dataset statistics v4.1 = lbs_v4 train+val (verbatim) + a test split absorbing every config from dac-research/extra_evals_v1 not already in v4. Five ZeroScrolls configs whose upstream corpora collide with lbs_v3.1 train/test (gov_report, qmsum, qasper, narrative_qa, musique) were dropped outright. All test rows have rubrics backfilled via the longbench_generate_rubric_for_imported.jinja template (same path as lbs_v3.1 / lbs_v4 Stage 1b).… See the full description on the dataset page: https://huggingface.co/datasets/dac-research/longbench_synthetic_v4_1.

<iframe src="https://silytics.si/embed/item/02feedcb-6472-4b55-9fde-0586306c7a4a" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.