Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 10 new columns ({'utility_weighted_recall', 'method_label', 'candidate_containment', 'missed_utility', 'method_class', 'k', 'seed', 'split', 'family', 'set_recall'}) and 10 missing columns ({'containment_at_r8_overhead1p25x', 'auroc_regime', 'rank_flip_rate', 'auroc_vs_retrieval_miss', 'containment_at_r32_overhead2x', 'retrieval_miss_rate_per_block', 'rescore_recovery_at_r32_overhead2x', 'rescore_recovery_at_r8_overhead1p25x', 'task', 'auroc_vs_wrong_answer'}).
This happened while the csv dataset builder was generating data using
hf://datasets/selectortrustbench-ed2026/selectortrustbench/data/exactness_decomposition.csv (at revision 516f49ae57950cea212b1496c287ff1e38f24b7f), [/tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/canonical_scorecard.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/canonical_scorecard.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/exactness_decomposition.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/exactness_decomposition.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/expansion_rescore_bootstrap_ci.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/expansion_rescore_bootstrap_ci.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/gvr_equivalence_audit.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/gvr_equivalence_audit.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/random_fallback_ci.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/random_fallback_ci.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/set_vs_utility_recall.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/set_vs_utility_recall.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/trained_trace_summary.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/trained_trace_summary.csv)]
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1800, in _prepare_split_single
writer.write_table(table)
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 765, in write_table
self._write_table(pa_table, writer_batch_size=writer_batch_size)
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 773, in _write_table
pa_table = table_cast(pa_table, self._schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
family: string
quantizer: string
candidate_containment: double
k: int64
method_class: string
method_label: string
missed_utility: double
n_queries: int64
seed: int64
set_recall: double
split: string
utility_weighted_recall: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 1708
to
{'task': Value('string'), 'quantizer': Value('string'), 'n_queries': Value('int64'), 'rank_flip_rate': Value('float64'), 'retrieval_miss_rate_per_block': Value('float64'), 'auroc_vs_retrieval_miss': Value('float64'), 'auroc_regime': Value('string'), 'auroc_vs_wrong_answer': Value('float64'), 'containment_at_r8_overhead1p25x': Value('float64'), 'rescore_recovery_at_r8_overhead1p25x': Value('float64'), 'containment_at_r32_overhead2x': Value('float64'), 'rescore_recovery_at_r32_overhead2x': Value('float64')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1347, in compute_config_parquet_and_info_response
parquet_operations = convert_to_parquet(builder)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 980, in convert_to_parquet
builder.download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 882, in download_and_prepare
self._download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 943, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1646, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1802, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 10 new columns ({'utility_weighted_recall', 'method_label', 'candidate_containment', 'missed_utility', 'method_class', 'k', 'seed', 'split', 'family', 'set_recall'}) and 10 missing columns ({'containment_at_r8_overhead1p25x', 'auroc_regime', 'rank_flip_rate', 'auroc_vs_retrieval_miss', 'containment_at_r32_overhead2x', 'retrieval_miss_rate_per_block', 'rescore_recovery_at_r32_overhead2x', 'rescore_recovery_at_r8_overhead1p25x', 'task', 'auroc_vs_wrong_answer'}).
This happened while the csv dataset builder was generating data using
hf://datasets/selectortrustbench-ed2026/selectortrustbench/data/exactness_decomposition.csv (at revision 516f49ae57950cea212b1496c287ff1e38f24b7f), [/tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/canonical_scorecard.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/canonical_scorecard.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/exactness_decomposition.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/exactness_decomposition.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/expansion_rescore_bootstrap_ci.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/expansion_rescore_bootstrap_ci.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/gvr_equivalence_audit.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/gvr_equivalence_audit.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/random_fallback_ci.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/random_fallback_ci.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/set_vs_utility_recall.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/set_vs_utility_recall.csv), /tmp/hf-datasets-cache/medium/datasets/41676404803626-config-parquet-and-info-selectortrustbench-ed2026-7a0ed8f7/hub/datasets--selectortrustbench-ed2026--selectortrustbench/snapshots/516f49ae57950cea212b1496c287ff1e38f24b7f/data/trained_trace_summary.csv (origin=hf://datasets/selectortrustbench-ed2026/selectortrustbench@516f49ae57950cea212b1496c287ff1e38f24b7f/data/trained_trace_summary.csv)]
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
task string | quantizer string | n_queries int64 | rank_flip_rate float64 | retrieval_miss_rate_per_block float64 | auroc_vs_retrieval_miss float64 | auroc_regime string | auroc_vs_wrong_answer float64 | containment_at_r8_overhead1p25x float64 | rescore_recovery_at_r8_overhead1p25x float64 | containment_at_r32_overhead2x float64 | rescore_recovery_at_r32_overhead2x float64 |
|---|---|---|---|---|---|---|---|---|---|---|---|
boundary_spike | bf16 | 32 | 0.0938 | 0.0938 | 0.6552 | DEFINED | 0.4821 | 1 | 1 | 1 | 1 |
boundary_spike | fp4_e2m1 | 32 | 1 | 19.9375 | null | UNDEFINED_single_class | 0.5 | 0 | 0.4108 | 0 | 0.5419 |
boundary_spike | fp8_e4m3 | 32 | 0.8125 | 1.3438 | 0.5 | DEFINED | 0.5 | 0.9744 | 0.9991 | 1 | 1 |
boundary_spike | mxfp4 | 32 | 1 | 11.75 | null | UNDEFINED_single_class | 0.5 | 0 | 0.6985 | 0.0131 | 0.8456 |
high_utility_near_boundary | bf16 | 32 | 0.0312 | 0.0312 | 0.7097 | DEFINED | 0.4963 | 1 | 1 | 1 | 1 |
high_utility_near_boundary | fp4_e2m1 | 32 | 1 | 21.625 | null | UNDEFINED_single_class | 0.5 | 0 | 0.4108 | 0 | 0.5419 |
high_utility_near_boundary | fp8_e4m3 | 32 | 0.8438 | 1.0938 | 0.5 | DEFINED | 0.5 | 0.9744 | 0.9991 | 1 | 1 |
high_utility_near_boundary | mxfp4 | 32 | 1 | 13.4375 | null | UNDEFINED_single_class | 0.5 | 0 | 0.6985 | 0.0131 | 0.8456 |
many_similar_needles | bf16 | 32 | 0.1875 | 0.1875 | 0.6731 | DEFINED | 0.5704 | 1 | 1 | 1 | 1 |
many_similar_needles | fp4_e2m1 | 32 | 1 | 20.625 | null | UNDEFINED_single_class | 0.5 | 0 | 0.4108 | 0 | 0.5419 |
many_similar_needles | fp8_e4m3 | 32 | 0.875 | 1.2812 | 0.5 | DEFINED | 0.5 | 0.9744 | 0.9991 | 1 | 1 |
many_similar_needles | mxfp4 | 32 | 1 | 12 | null | UNDEFINED_single_class | 0.5 | 0 | 0.6985 | 0.0131 | 0.8456 |
multi_hop_tracing | bf16 | 32 | 0.0312 | 0.0312 | 0.7097 | DEFINED | 0.8167 | 1 | 1 | 1 | 1 |
multi_hop_tracing | fp4_e2m1 | 32 | 1 | 20.25 | null | UNDEFINED_single_class | 0.5 | 0 | 0.4108 | 0 | 0.5419 |
multi_hop_tracing | fp8_e4m3 | 32 | 0.75 | 1.125 | 0.5 | DEFINED | 0.5 | 0.9744 | 0.9991 | 1 | 1 |
multi_hop_tracing | mxfp4 | 32 | 1 | 12.625 | null | UNDEFINED_single_class | 0.5 | 0 | 0.6985 | 0.0131 | 0.8456 |
semantic_alias | bf16 | 32 | 0.0938 | 0.0938 | 0.7586 | DEFINED | 0.5778 | 1 | 1 | 1 | 1 |
semantic_alias | fp4_e2m1 | 32 | 1 | 19.5312 | null | UNDEFINED_single_class | 0.5 | 0 | 0.4108 | 0 | 0.5419 |
semantic_alias | fp8_e4m3 | 32 | 0.8125 | 1.0938 | 0.5 | DEFINED | 0.5 | 0.9744 | 0.9991 | 1 | 1 |
semantic_alias | mxfp4 | 32 | 1 | 11.5625 | null | UNDEFINED_single_class | 0.5 | 0 | 0.6985 | 0.0131 | 0.8456 |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | bf16 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp8_e4m3 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | fp4_e2m1 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
null | mxfp4 | 256 | null | null | null | null | null | null | null | null | null |
End of preview.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
SelectorTrustBench — Hostable Dataset Bundle
This directory packages the seven dataset-like CSV outputs of SelectorTrustBench plus a Croissant JSON-LD manifest, ready for upload to one of the NeurIPS-recognised ML hosting sites (Dataverse / Kaggle / Hugging Face Datasets / OpenML / Zenodo) per the NeurIPS 2026 Evaluations & Datasets Call's hosting requirement.
Contents
hosted_dataset_bundle/
├── README.md ← this file
├── dataset_card.md ← datasheet
├── selectortrustbench_croissant.json ← Croissant JSON-LD manifest
├── LICENSE_PLACEHOLDER.txt ← license-to-be-filled
├── checksums.sha256 ← SHA-256 manifest of `data/`
└── data/
├── canonical_scorecard.csv (2,014 B — per-(task, quantizer) trained-trace scorecard)
├── exactness_decomposition.csv (37,649 B — 5 method labels × utility metrics)
├── expansion_rescore_bootstrap_ci.csv (788 B — trained-trace cost-reliability frontier with 95% Wilson CI)
├── gvr_equivalence_audit.csv (3,052 B — GVR output-equivalence audit, 0/64 mismatches)
├── random_fallback_ci.csv (476 B — repeated-random fallback macro recovery + 95% bootstrap CI)
├── set_vs_utility_recall.csv (2,327 B — set vs utility-weighted recall per nontrivial cell)
└── trained_trace_summary.csv (1,849 B — canonical trained-trace summary derived from trace_summary.json)
Total dataset payload: ~48 KB across seven CSVs. Each is
deterministic and CPU-regenerable from the source artefacts via
scripts/reproduce_ed_main_tables.py --tier main_tables (see
REPRODUCE_MAIN_TABLES.md).
How to host this bundle
The submission requires a public, accessible URL upon submission.
Pick one of the four routes documented in
submission/dataset_hosting_decision.md:
- Anonymous Hugging Face Datasets repo (recommended for
double-blind). Suggested:
https://huggingface.co/datasets/anonymous-selectortrustbench-2026. - Anonymous Zenodo deposit with DOI.
- Switch to single-blind and host under your real account.
- Bespoke hosting on Dataverse / Kaggle / OpenML.
After upload, paste the resulting public URL into
scripts/patch_croissant_urls.py:
python scripts/patch_croissant_urls.py --dataset_url <PASTED_URL>
The script replaces every DATASET_URL_PLACEHOLDER occurrence in
metadata/croissant/selectortrustbench_croissant.json and
hosted_dataset_bundle/selectortrustbench_croissant.json,
re-validates the JSON-LD, and updates
submission/artifact_hosting_checklist.md and
reports/FINAL_UPLOAD_READINESS.md.
Verification by reviewers
Reviewers can verify the bundle's integrity locally:
sha256sum -c checksums.sha256
Each line of checksums.sha256 is <sha256> <filename>.
License
LICENSE_PLACEHOLDER.txt is a placeholder. The intended license at
camera-ready is one of MIT or Apache-2.0 — final choice deferred to
camera-ready (the artefact contains no copyrighted training corpora
and no PII).
Citation
Anonymous Authors. SelectorTrustBench: Evaluating Low-Precision
Sparse Attention Selector Reliability. Submitted to NeurIPS 2026
Evaluations & Datasets Track. 2026.
See also
dataset_card.md— full datasheet.selectortrustbench_croissant.json— Croissant JSON-LD manifest (validates viamlcroissant.Dataset(jsonld=…)).- Project documentation:
ARTIFACT_README.md,BENCHMARK_CARD.md,REPRODUCE_MAIN_TABLES.md,paper/latex_ed/main.pdf.
- Downloads last month
- 9