Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: CastError
Message: Couldn't cast
config_a: string
config_b: string
n_pairs: int64
mean_delta_decode_tps: double
ci95_delta_low: double
ci95_delta_high: double
bootstrap_delta_low: double
bootstrap_delta_high: double
mean_delta_pct: double
bootstrap_pct_low: double
bootstrap_pct_high: double
paired_t_stat: double
paired_t_p: double
wilcoxon_z: double
wilcoxon_p: double
cohens_d: double
holm_wilcoxon_p: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 2424
to
{'prompt_id': Value('string'), 'source': Value('string'), 'attention_decode_tps': Value('float64'), 'saliency_decode_tps': Value('float64'), 'delta_decode_tps': Value('float64'), 'delta_pct': Value('float64'), 'attention_predicted_n': Value('int64'), 'saliency_predicted_n': Value('int64')}
because column names don't match
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 764, in write_table
self.write_rows_on_file() # in case there are buffered rows to write first
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 663, in write_rows_on_file
self._write_table(table)
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
config_a: string
config_b: string
n_pairs: int64
mean_delta_decode_tps: double
ci95_delta_low: double
ci95_delta_high: double
bootstrap_delta_low: double
bootstrap_delta_high: double
mean_delta_pct: double
bootstrap_pct_low: double
bootstrap_pct_high: double
paired_t_stat: double
paired_t_p: double
wilcoxon_z: double
wilcoxon_p: double
cohens_d: double
holm_wilcoxon_p: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 2424
to
{'prompt_id': Value('string'), 'source': Value('string'), 'attention_decode_tps': Value('float64'), 'saliency_decode_tps': Value('float64'), 'delta_decode_tps': Value('float64'), 'delta_pct': Value('float64'), 'attention_predicted_n': Value('int64'), 'saliency_predicted_n': Value('int64')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
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 17 new columns ({'wilcoxon_p', 'paired_t_p', 'mean_delta_decode_tps', 'config_a', 'ci95_delta_high', 'paired_t_stat', 'ci95_delta_low', 'mean_delta_pct', 'bootstrap_pct_high', 'cohens_d', 'holm_wilcoxon_p', 'bootstrap_delta_low', 'wilcoxon_z', 'bootstrap_delta_high', 'bootstrap_pct_low', 'n_pairs', 'config_b'}) and 8 missing columns ({'prompt_id', 'attention_predicted_n', 'attention_decode_tps', 'saliency_decode_tps', 'source', 'saliency_predicted_n', 'delta_pct', 'delta_decode_tps'}).
This happened while the csv dataset builder was generating data using
hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100/results.csv (at revision e08bd7700e3c6f88e59572a9e46bd79fe5b14e76), [/tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/outliers.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/outliers.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/paired_comparisons.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/paired_comparisons.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/results.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/results.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/source_breakdown.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/source_breakdown.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/statistical_measures.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/statistical_measures.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/summary.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/summary.csv), /tmp/hf-datasets-cache/medium/datasets/30108620612057-config-parquet-and-info-sjakek-qwen36-q4km-salien-1b001f51/hub/datasets--sjakek--qwen36-q4km-saliency-vs-attention-residency-100/snapshots/e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/tweet_summary_table.csv (origin=hf://datasets/sjakek/qwen36-q4km-saliency-vs-attention-residency-100@e08bd7700e3c6f88e59572a9e46bd79fe5b14e76/tweet_summary_table.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)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1821, in _prepare_split_single
num_examples, num_bytes = writer.finalize()
^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 781, in finalize
self.write_rows_on_file()
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 663, in write_rows_on_file
self._write_table(table)
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
config_a: string
config_b: string
n_pairs: int64
mean_delta_decode_tps: double
ci95_delta_low: double
ci95_delta_high: double
bootstrap_delta_low: double
bootstrap_delta_high: double
mean_delta_pct: double
bootstrap_pct_low: double
bootstrap_pct_high: double
paired_t_stat: double
paired_t_p: double
wilcoxon_z: double
wilcoxon_p: double
cohens_d: double
holm_wilcoxon_p: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 2424
to
{'prompt_id': Value('string'), 'source': Value('string'), 'attention_decode_tps': Value('float64'), 'saliency_decode_tps': Value('float64'), 'delta_decode_tps': Value('float64'), 'delta_pct': Value('float64'), 'attention_predicted_n': Value('int64'), 'saliency_predicted_n': Value('int64')}
because column names don't match
The above exception was the direct cause of the following exception:
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 1832, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
prompt_id string | source string | attention_decode_tps float64 | saliency_decode_tps float64 | delta_decode_tps float64 | delta_pct float64 | attention_predicted_n int64 | saliency_predicted_n int64 |
|---|---|---|---|---|---|---|---|
tbench_feal-linear-cryptanalysis | terminal_bench_hard | 54.359868 | 74.270198 | 19.910329 | 36.62689 | 4,096 | 4,096 |
tbench_neuron-to-jaxley-conversion | terminal_bench_hard | 60.682189 | 74.924248 | 14.242059 | 23.469917 | 773 | 4,096 |
tbench_play-zork-easy | terminal_bench_hard | 62.291049 | 74.610291 | 12.319242 | 19.776906 | 4,096 | 4,096 |
tbench_swe-bench-astropy-2 | terminal_bench_hard | 67.527052 | 79.960278 | 12.433226 | 18.412215 | 4,096 | 4,096 |
tbench_form-filling | terminal_bench_hard | 63.324006 | 74.965152 | 11.641145 | 18.383463 | 4,096 | 4,096 |
tbench_gpt2-codegolf | terminal_bench_hard | 63.087785 | 73.300531 | 10.212745 | 16.18815 | 4,096 | 4,096 |
scicode_33_33.2 | scicode | 65.140607 | 75.551355 | 10.410748 | 15.981963 | 4,096 | 4,096 |
tbench_prove-plus-comm | terminal_bench_hard | 63.035539 | 72.375549 | 9.340011 | 14.817055 | 898 | 748 |
scicode_64_64.1 | scicode | 61.889857 | 71.000904 | 9.111047 | 14.72139 | 224 | 833 |
tbench_make-mips-interpreter | terminal_bench_hard | 67.270221 | 77.058404 | 9.788183 | 14.550544 | 4,096 | 4,096 |
tbench_run-pdp11-code | terminal_bench_hard | 61.93049 | 70.728281 | 8.797791 | 14.205911 | 1,835 | 2,148 |
scicode_58_58.5 | scicode | 62.789777 | 71.544493 | 8.754716 | 13.9429 | 4,096 | 4,096 |
tbench_pytorch-model-cli | terminal_bench_hard | 65.274788 | 74.359137 | 9.084349 | 13.917086 | 4,096 | 3,724 |
tbench_extract-moves-from-video | terminal_bench_hard | 63.634148 | 72.395388 | 8.76124 | 13.768142 | 808 | 4,096 |
scicode_59_59.5 | scicode | 65.556573 | 74.495346 | 8.938773 | 13.635204 | 4,096 | 1,544 |
tbench_git-multibranch | terminal_bench_hard | 65.169339 | 73.910992 | 8.741652 | 13.41375 | 1,516 | 4,096 |
scicode_57_57.2 | scicode | 65.736932 | 74.48295 | 8.746018 | 13.304572 | 994 | 2,340 |
scicode_70_70.2 | scicode | 68.718998 | 77.785072 | 9.066074 | 13.192966 | 2,157 | 4,096 |
scicode_57_57.5 | scicode | 63.441431 | 71.799208 | 8.357777 | 13.174005 | 4,096 | 4,096 |
tbench_movie-helper | terminal_bench_hard | 61.221018 | 69.151346 | 7.930328 | 12.953604 | 4,096 | 4,096 |
null | null | null | null | null | null | null | null |
scicode_10_10.10 | scicode | null | null | null | null | null | null |
scicode_11_11.10 | scicode | null | null | null | null | null | null |
scicode_11_11.11 | scicode | null | null | null | null | null | null |
scicode_11_11.9 | scicode | null | null | null | null | null | null |
scicode_13_13.13 | scicode | null | null | null | null | null | null |
scicode_14_14.2 | scicode | null | null | null | null | null | null |
scicode_20_20.2 | scicode | null | null | null | null | null | null |
scicode_30_30.1 | scicode | null | null | null | null | null | null |
scicode_32_32.3 | scicode | null | null | null | null | null | null |
scicode_33_33.1 | scicode | null | null | null | null | null | null |
scicode_33_33.2 | scicode | null | null | null | null | null | null |
scicode_34_34.3 | scicode | null | null | null | null | null | null |
scicode_36_36.1 | scicode | null | null | null | null | null | null |
scicode_37_37.2 | scicode | null | null | null | null | null | null |
scicode_38_38.1 | scicode | null | null | null | null | null | null |
scicode_42_42.2 | scicode | null | null | null | null | null | null |
scicode_42_42.3 | scicode | null | null | null | null | null | null |
scicode_45_45.1 | scicode | null | null | null | null | null | null |
scicode_46_46.2 | scicode | null | null | null | null | null | null |
scicode_48_48.4 | scicode | null | null | null | null | null | null |
scicode_53_53.2 | scicode | null | null | null | null | null | null |
scicode_55_55.3 | scicode | null | null | null | null | null | null |
scicode_57_57.2 | scicode | null | null | null | null | null | null |
scicode_57_57.5 | scicode | null | null | null | null | null | null |
scicode_58_58.5 | scicode | null | null | null | null | null | null |
scicode_59_59.1 | scicode | null | null | null | null | null | null |
scicode_59_59.4 | scicode | null | null | null | null | null | null |
scicode_59_59.5 | scicode | null | null | null | null | null | null |
scicode_60_60.1 | scicode | null | null | null | null | null | null |
scicode_60_60.2 | scicode | null | null | null | null | null | null |
scicode_61_61.1 | scicode | null | null | null | null | null | null |
scicode_63_63.1 | scicode | null | null | null | null | null | null |
scicode_64_64.1 | scicode | null | null | null | null | null | null |
scicode_64_64.2 | scicode | null | null | null | null | null | null |
scicode_64_64.5 | scicode | null | null | null | null | null | null |
scicode_65_65.4 | scicode | null | null | null | null | null | null |
scicode_66_66.1 | scicode | null | null | null | null | null | null |
scicode_67_67.5 | scicode | null | null | null | null | null | null |
scicode_68_68.6 | scicode | null | null | null | null | null | null |
scicode_68_68.7 | scicode | null | null | null | null | null | null |
scicode_69_69.6 | scicode | null | null | null | null | null | null |
scicode_70_70.2 | scicode | null | null | null | null | null | null |
scicode_70_70.4 | scicode | null | null | null | null | null | null |
scicode_70_70.7 | scicode | null | null | null | null | null | null |
scicode_71_71.8 | scicode | null | null | null | null | null | null |
scicode_72_72.3 | scicode | null | null | null | null | null | null |
scicode_72_72.6 | scicode | null | null | null | null | null | null |
scicode_72_72.7 | scicode | null | null | null | null | null | null |
scicode_72_72.9 | scicode | null | null | null | null | null | null |
scicode_73_73.3 | scicode | null | null | null | null | null | null |
scicode_73_73.4 | scicode | null | null | null | null | null | null |
scicode_76_76.1 | scicode | null | null | null | null | null | null |
scicode_76_76.4 | scicode | null | null | null | null | null | null |
scicode_77_77.7 | scicode | null | null | null | null | null | null |
scicode_78_78.1 | scicode | null | null | null | null | null | null |
scicode_8_8.1 | scicode | null | null | null | null | null | null |
tbench_aimo-airline-departures | terminal_bench_hard | null | null | null | null | null | null |
tbench_blind-maze-explorer-5x5 | terminal_bench_hard | null | null | null | null | null | null |
tbench_cartpole-rl-training | terminal_bench_hard | null | null | null | null | null | null |
tbench_chem-property-targeting | terminal_bench_hard | null | null | null | null | null | null |
tbench_chem-rf | terminal_bench_hard | null | null | null | null | null | null |
tbench_circuit-fibsqrt | terminal_bench_hard | null | null | null | null | null | null |
tbench_cobol-modernization | terminal_bench_hard | null | null | null | null | null | null |
tbench_configure-git-webserver | terminal_bench_hard | null | null | null | null | null | null |
tbench_cross-entropy-method | terminal_bench_hard | null | null | null | null | null | null |
tbench_extract-moves-from-video | terminal_bench_hard | null | null | null | null | null | null |
tbench_feal-differential-cryptanalysis | terminal_bench_hard | null | null | null | null | null | null |
tbench_feal-linear-cryptanalysis | terminal_bench_hard | null | null | null | null | null | null |
tbench_form-filling | terminal_bench_hard | null | null | null | null | null | null |
tbench_git-multibranch | terminal_bench_hard | null | null | null | null | null | null |
tbench_gpt2-codegolf | terminal_bench_hard | null | null | null | null | null | null |
tbench_install-windows-xp | terminal_bench_hard | null | null | null | null | null | null |
tbench_make-doom-for-mips | terminal_bench_hard | null | null | null | null | null | null |
tbench_make-mips-interpreter | terminal_bench_hard | null | null | null | null | null | null |
tbench_model-extraction-relu-logits | terminal_bench_hard | null | null | null | null | null | null |
tbench_movie-helper | terminal_bench_hard | null | null | null | null | null | null |
tbench_neuron-to-jaxley-conversion | terminal_bench_hard | null | null | null | null | null | null |
tbench_oom | terminal_bench_hard | null | null | null | null | null | null |
tbench_organization-json-generator | terminal_bench_hard | null | null | null | null | null | null |
End of preview.
Qwen3.6 Q4_K_M Saliency vs Attention Residency
This dataset package contains a paired 100-prompt throughput replay comparing
the original attention-spaced MoE layer residency policy against the
saliency-selected b11 hot-layer policy for
Qwen3.6-35B-A3B-UD-Q4_K_M.gguf.
The prompt file is an exact replay of qwen36-moe-layer-residency-20260520-205703: 56 SciCode
subproblems and 44 Artificial Analysis Terminal-Bench-Hard prompts. The run
uses 64k context, q8_0/q8_0 KV cache, Flash Attention, and MTP with
--spec-type mtp --spec-draft-n-max 2.
See analysis_q4km_saliency_vs_attention.html for the statistical analysis
and tweet_summary_table.md for a compact sharing table.
- Downloads last month
- 115