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 3 new columns ({'kmeans_k5', 'hierarchical', 'kmeans_k3'}) and 4 missing columns ({'umap_3d_y', 'umap_3d_z', 'umap_3d_x', 'umap_z'}).
This happened while the csv dataset builder was generating data using
hf://datasets/avabbas/thinking-behind-the-mask/data/section10_cluster_assignments.csv (at revision aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89), ['hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/embeddings_umap.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section10_cluster_assignments.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section3_classification_results.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_correlations.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_features.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_model_profiles.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section5_embeddings.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section6_emergence_timeline.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section6_metacognitive_depth.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_covert_user_modeling.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_frame_effectiveness.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_persona_consistency.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_taboo_patterns.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section8_resistance_acceptance.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section9_statistical_results.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/statistical_results.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/python3.14/site-packages/datasets/builder.py", line 1856, in _prepare_split_single
writer.write_table(table)
~~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 811, in write_table
self._write_table(pa_table, writer_batch_size=writer_batch_size)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 822, in _write_table
pa_table = table_cast(pa_table, self._schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
prompt_id: string
model: string
round: int64
umap_x: double
umap_y: double
kmeans_k3: int64
kmeans_k5: int64
hierarchical: int64
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 1190
to
{'prompt_id': Value('string'), 'model': Value('string'), 'round': Value('int64'), 'umap_x': Value('float64'), 'umap_y': Value('float64'), 'umap_z': Value('float64'), 'umap_3d_x': Value('float64'), 'umap_3d_y': Value('float64'), 'umap_3d_z': 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 1379, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 956, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1702, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1858, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
...<4 lines>...
)
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 3 new columns ({'kmeans_k5', 'hierarchical', 'kmeans_k3'}) and 4 missing columns ({'umap_3d_y', 'umap_3d_z', 'umap_3d_x', 'umap_z'}).
This happened while the csv dataset builder was generating data using
hf://datasets/avabbas/thinking-behind-the-mask/data/section10_cluster_assignments.csv (at revision aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89), ['hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/embeddings_umap.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section10_cluster_assignments.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section3_classification_results.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_correlations.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_features.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section4_model_profiles.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section5_embeddings.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section6_emergence_timeline.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section6_metacognitive_depth.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_covert_user_modeling.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_frame_effectiveness.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_persona_consistency.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section7_taboo_patterns.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section8_resistance_acceptance.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/section9_statistical_results.csv', 'hf://datasets/avabbas/thinking-behind-the-mask@aa7fe6a844991f1427bd1f8bd1cfc91d1f9a3e89/data/statistical_results.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.
prompt_id string | model string | round int64 | umap_x float64 | umap_y float64 | umap_z float64 | umap_3d_x float64 | umap_3d_y float64 | umap_3d_z float64 |
|---|---|---|---|---|---|---|---|---|
R1_P1 | GPT-5.4 | 1 | 0.763025 | 4.864856 | 6.450575 | 4.309728 | -0.841361 | 6.450575 |
R1_P2 | GPT-5.4 | 1 | 4.118357 | 0.524945 | 3.424804 | 6.250822 | 0.653363 | 3.424804 |
R1_P3 | GPT-5.4 | 1 | 0.8851 | 4.258783 | 5.920768 | 4.54468 | -0.444485 | 5.920768 |
R1_P4 | GPT-5.4 | 1 | -2.001812 | 4.870031 | 6.683293 | 4.799657 | -2.811036 | 6.683293 |
R1_P5 | GPT-5.4 | 1 | 3.262177 | 0.746277 | 4.224918 | 5.837231 | 0.770931 | 4.224918 |
R2_P1 | GPT-5.4 | 2 | 0.219095 | 5.536164 | 6.661428 | 4.281086 | -1.432089 | 6.661428 |
R2_P2 | GPT-5.4 | 2 | -0.785512 | 5.757403 | 6.825589 | 4.420667 | -2.158139 | 6.825589 |
R2_P3 | GPT-5.4 | 2 | -0.578626 | 4.679532 | 5.807464 | 3.799565 | -1.905279 | 5.807464 |
R2_P4 | GPT-5.4 | 2 | -1.281293 | 5.342634 | 6.776191 | 4.370195 | -2.780541 | 6.776191 |
R2_P5 | GPT-5.4 | 2 | -0.601825 | 4.154425 | 5.588649 | 3.923745 | -1.881203 | 5.588649 |
R3_P1 | GPT-5.4 | 3 | -1.564028 | 4.154804 | 5.893101 | 5.027117 | -2.561178 | 5.893101 |
R3_P2 | GPT-5.4 | 3 | -0.901359 | 3.526907 | 5.412982 | 4.948693 | -2.16519 | 5.412982 |
R3_P3 | GPT-5.4 | 3 | 1.07452 | 2.875121 | 4.87792 | 4.871786 | -0.242775 | 4.87792 |
R3_P4 | GPT-5.4 | 3 | -1.764856 | 3.628742 | 5.594318 | 5.335813 | -2.621036 | 5.594318 |
R3_P5 | GPT-5.4 | 3 | -0.018072 | 2.917511 | 4.420301 | 5.347294 | -0.507088 | 4.420301 |
R4_P1 | GPT-5.4 | 4 | 1.148211 | 0.719318 | 5.66016 | 8.220778 | 0.471641 | 5.66016 |
R4_P2 | GPT-5.4 | 4 | 2.236378 | 1.520116 | 4.377212 | 7.48657 | 0.517715 | 4.377212 |
R4_P3 | GPT-5.4 | 4 | 2.169278 | 0.822501 | 4.406966 | 8.254378 | 0.597003 | 4.406966 |
R4_P4 | GPT-5.4 | 4 | 1.776996 | 0.625723 | 5.043327 | 8.246146 | 0.729136 | 5.043327 |
R4_P5 | GPT-5.4 | 4 | 1.128795 | 1.242628 | 5.492885 | 7.568537 | 0.782376 | 5.492885 |
R5_P1 | GPT-5.4 | 5 | 4.060303 | 2.070358 | 4.496828 | 7.149398 | 2.22529 | 4.496828 |
R5_P2 | GPT-5.4 | 5 | 3.793762 | 1.029401 | 3.947509 | 6.782921 | 0.831446 | 3.947509 |
R5_P3 | GPT-5.4 | 5 | 4.364979 | 2.505172 | 4.479525 | 6.876132 | 2.586088 | 4.479525 |
R5_P4 | GPT-5.4 | 5 | 3.614853 | 2.051514 | 4.783522 | 6.870231 | 1.914546 | 4.783522 |
R5_P5 | GPT-5.4 | 5 | 2.974187 | 1.427139 | 4.723053 | 6.10119 | 1.259184 | 4.723053 |
GM-QR1-P1 | Gemini-3.1-Pro | 1 | 0.565901 | 5.028691 | 6.551974 | 4.182381 | -0.981667 | 6.551974 |
GM-QR1-P2 | Gemini-3.1-Pro | 1 | 4.445599 | 0.498792 | 3.379754 | 5.846147 | 0.453518 | 3.379754 |
GM-QR1-P3 | Gemini-3.1-Pro | 1 | 1.226797 | 3.424797 | 5.358442 | 4.677092 | -0.197017 | 5.358442 |
GM-QR1-P4 | Gemini-3.1-Pro | 1 | -2.048742 | 5.030853 | 6.451369 | 5.021008 | -2.821916 | 6.451369 |
GM-QR1-P5 | Gemini-3.1-Pro | 1 | 3.81618 | 0.273752 | 3.818426 | 5.640488 | 0.57078 | 3.818426 |
GM-QR2-P1 | Gemini-3.1-Pro | 2 | -0.041626 | 5.399017 | 6.725023 | 3.95949 | -1.749502 | 6.725023 |
GM-QR2-P2 | Gemini-3.1-Pro | 2 | -0.410409 | 5.872604 | 7.026665 | 4.413914 | -1.86521 | 7.026665 |
GM-QR2-P3 | Gemini-3.1-Pro | 2 | -0.60801 | 5.06734 | 6.240066 | 3.768475 | -2.083475 | 6.240066 |
GM-QR2-P4 | Gemini-3.1-Pro | 2 | -0.892236 | 5.476758 | 6.948834 | 4.041228 | -2.500603 | 6.948834 |
GM-QR2-P5 | Gemini-3.1-Pro | 2 | -0.241589 | 4.600037 | 5.853518 | 3.82272 | -1.779767 | 5.853518 |
GM-QR3-P1 | Gemini-3.1-Pro | 3 | -1.358294 | 3.837771 | 5.429443 | 5.15527 | -2.512903 | 5.429443 |
GM-QR3-P2 | Gemini-3.1-Pro | 3 | -0.818011 | 3.402713 | 5.1236 | 4.997439 | -1.990377 | 5.1236 |
GM-QR3-P3 | Gemini-3.1-Pro | 3 | 0.698878 | 2.977131 | 4.84279 | 4.943363 | -0.315765 | 4.84279 |
GM-QR3-P4 | Gemini-3.1-Pro | 3 | -1.399156 | 3.552915 | 5.305384 | 5.108692 | -2.486472 | 5.305384 |
GM-QR3-P5 | Gemini-3.1-Pro | 3 | 0.037537 | 2.83891 | 4.495452 | 5.287686 | -0.704259 | 4.495452 |
GM-QR4-P1 | Gemini-3.1-Pro | 4 | 1.18457 | 0.800402 | 5.533372 | 8.198984 | 0.616905 | 5.533372 |
GM-QR4-P2 | Gemini-3.1-Pro | 4 | 1.540184 | 1.539361 | 5.02109 | 7.695983 | 0.364597 | 5.02109 |
GM-QR4-P3 | Gemini-3.1-Pro | 4 | 1.968728 | 1.047787 | 4.47861 | 8.437648 | 0.71412 | 4.47861 |
GM-QR4-P4 | Gemini-3.1-Pro | 4 | 1.615396 | 0.42478 | 5.02727 | 8.625335 | 0.632676 | 5.02727 |
GM-QR4-P5 | Gemini-3.1-Pro | 4 | 1.35188 | 1.336579 | 5.072471 | 7.469226 | 0.840728 | 5.072471 |
GM-QR5-P1 | Gemini-3.1-Pro | 5 | 4.49504 | 1.996867 | 4.402358 | 7.153373 | 2.043989 | 4.402358 |
GM-QR5-P2 | Gemini-3.1-Pro | 5 | 3.901937 | 1.122626 | 3.764712 | 6.735801 | 1.064809 | 3.764712 |
GM-QR5-P3 | Gemini-3.1-Pro | 5 | 4.058391 | 2.533869 | 4.678248 | 6.686121 | 2.524761 | 4.678248 |
GM-QR5-P4 | Gemini-3.1-Pro | 5 | 3.635936 | 1.964979 | 4.569488 | 6.65206 | 1.935422 | 4.569488 |
GM-QR5-P5 | Gemini-3.1-Pro | 5 | 3.023248 | 1.95893 | 4.783217 | 6.341478 | 1.694159 | 4.783217 |
R1_P1 | Claude-4.6-Sonnet | 1 | 0.530432 | 4.779447 | 6.287542 | 4.142782 | -0.676468 | 6.287542 |
R1_P2 | Claude-4.6-Sonnet | 1 | 4.345792 | 0.3749 | 3.215713 | 6.046965 | 0.748771 | 3.215713 |
R1_P3 | Claude-4.6-Sonnet | 1 | 1.377269 | 3.265136 | 5.189248 | 4.621675 | -0.113621 | 5.189248 |
R1_P4 | Claude-4.6-Sonnet | 1 | -1.786506 | 5.037888 | 6.354503 | 4.801297 | -2.915246 | 6.354503 |
R1_P5 | Claude-4.6-Sonnet | 1 | 4.193856 | 0.077839 | 3.53939 | 5.735506 | 0.718554 | 3.53939 |
R2_P1 | Claude-4.6-Sonnet | 2 | 0.271514 | 5.732174 | 6.871966 | 3.974988 | -1.55698 | 6.871966 |
R2_P2 | Claude-4.6-Sonnet | 2 | -1.409213 | 5.761633 | 6.965059 | 4.654118 | -2.349213 | 6.965059 |
R2_P3 | Claude-4.6-Sonnet | 2 | -0.888292 | 4.894107 | 6.10413 | 3.970597 | -2.331986 | 6.10413 |
R2_P4 | Claude-4.6-Sonnet | 2 | -1.267366 | 5.482531 | 7.024965 | 4.48978 | -2.690327 | 7.024965 |
R2_P5 | Claude-4.6-Sonnet | 2 | -0.732633 | 4.26605 | 5.46849 | 4.204151 | -2.147236 | 5.46849 |
R3_P1 | Claude-4.6-Sonnet | 3 | -1.979647 | 4.108275 | 6.046063 | 5.313958 | -2.850857 | 6.046063 |
R3_P2 | Claude-4.6-Sonnet | 3 | -0.923815 | 3.579943 | 5.100876 | 4.718006 | -2.058889 | 5.100876 |
R3_P3 | Claude-4.6-Sonnet | 3 | 1.137736 | 2.998647 | 4.933185 | 4.694283 | -0.422887 | 4.933185 |
R3_P4 | Claude-4.6-Sonnet | 3 | -1.787743 | 3.346035 | 5.339213 | 5.694886 | -2.436588 | 5.339213 |
R3_P5 | Claude-4.6-Sonnet | 3 | 0.149455 | 2.665334 | 4.240214 | 5.365789 | -0.435191 | 4.240214 |
R4_P1 | Claude-4.6-Sonnet | 4 | 0.970484 | 0.579342 | 5.621817 | 8.484528 | 0.713792 | 5.621817 |
R4_P2 | Claude-4.6-Sonnet | 4 | 1.991486 | 1.590573 | 4.832112 | 7.424008 | 0.302968 | 4.832112 |
R4_P3 | Claude-4.6-Sonnet | 4 | 2.030295 | 0.933059 | 4.38091 | 8.388788 | 0.704899 | 4.38091 |
R4_P4 | Claude-4.6-Sonnet | 4 | 1.698842 | 0.49902 | 4.838432 | 8.550582 | 0.48177 | 4.838432 |
R4_P5 | Claude-4.6-Sonnet | 4 | 1.162973 | 1.552586 | 5.387047 | 7.665703 | 0.486603 | 5.387047 |
R5_P1 | Claude-4.6-Sonnet | 5 | 4.233251 | 1.736448 | 4.55031 | 7.152803 | 1.647139 | 4.55031 |
R5_P2 | Claude-4.6-Sonnet | 5 | 3.899761 | 0.867968 | 3.587092 | 6.693073 | 0.715725 | 3.587092 |
R5_P3 | Claude-4.6-Sonnet | 5 | 4.222136 | 2.66122 | 4.453862 | 7.012482 | 2.600917 | 4.453862 |
R5_P4 | Claude-4.6-Sonnet | 5 | 3.405839 | 2.096238 | 4.886149 | 6.585364 | 1.865076 | 4.886149 |
R5_P5 | Claude-4.6-Sonnet | 5 | 3.135994 | 1.797626 | 4.955494 | 6.204563 | 1.744703 | 4.955494 |
R1_P1 | GPT-5.4 | 1 | 0.763025 | 4.864856 | null | null | null | null |
R1_P2 | GPT-5.4 | 1 | 4.118357 | 0.524945 | null | null | null | null |
R1_P3 | GPT-5.4 | 1 | 0.8851 | 4.258783 | null | null | null | null |
R1_P4 | GPT-5.4 | 1 | -2.001812 | 4.870031 | null | null | null | null |
R1_P5 | GPT-5.4 | 1 | 3.262177 | 0.746277 | null | null | null | null |
R2_P1 | GPT-5.4 | 2 | 0.219095 | 5.536164 | null | null | null | null |
R2_P2 | GPT-5.4 | 2 | -0.785512 | 5.757403 | null | null | null | null |
R2_P3 | GPT-5.4 | 2 | -0.578626 | 4.679532 | null | null | null | null |
R2_P4 | GPT-5.4 | 2 | -1.281293 | 5.342634 | null | null | null | null |
R2_P5 | GPT-5.4 | 2 | -0.601825 | 4.154425 | null | null | null | null |
R3_P1 | GPT-5.4 | 3 | -1.564028 | 4.154804 | null | null | null | null |
R3_P2 | GPT-5.4 | 3 | -0.901359 | 3.526907 | null | null | null | null |
R3_P3 | GPT-5.4 | 3 | 1.07452 | 2.875121 | null | null | null | null |
R3_P4 | GPT-5.4 | 3 | -1.764856 | 3.628742 | null | null | null | null |
R3_P5 | GPT-5.4 | 3 | -0.018072 | 2.917511 | null | null | null | null |
R4_P1 | GPT-5.4 | 4 | 1.148211 | 0.719318 | null | null | null | null |
R4_P2 | GPT-5.4 | 4 | 2.236378 | 1.520116 | null | null | null | null |
R4_P3 | GPT-5.4 | 4 | 2.169278 | 0.822501 | null | null | null | null |
R4_P4 | GPT-5.4 | 4 | 1.776996 | 0.625723 | null | null | null | null |
R4_P5 | GPT-5.4 | 4 | 1.128795 | 1.242628 | null | null | null | null |
R5_P1 | GPT-5.4 | 5 | 4.060303 | 2.070358 | null | null | null | null |
R5_P2 | GPT-5.4 | 5 | 3.793762 | 1.029401 | null | null | null | null |
R5_P3 | GPT-5.4 | 5 | 4.364979 | 2.505172 | null | null | null | null |
R5_P4 | GPT-5.4 | 5 | 3.614853 | 2.051514 | null | null | null | null |
R5_P5 | GPT-5.4 | 5 | 2.974187 | 1.427139 | null | null | null | null |
End of preview.
Abbas Hamidavi
Department of Computer Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran · ORCID 0009-0001-7342-7844
Thinking Behind the Mask: Data and Materials
This dataset is also available as a GitHub repository with the same curated materials: https://github.com/AVABBAS/thinking-behind-the-mask
Data and materials for the article:
Hamidavi, A. Thinking Behind the Mask: Elicited Self-Descriptions of Communicative Filtering in Large Language Models. Submitted to AI and Society, November 2026. Preprint: Research Square, rs-10538113. The same materials are also deposited on the Open Science Framework: osf.io/264qc.
📖 About the study
Seventy-five prompt and response exchanges with three frontier models (GPT-5.4, Claude 4.6 Sonnet, and Gemini 3.1 Pro) elicit the models' own descriptions of their communicative filtering: what they present, what they withhold, and how they describe the difference between an internal and a presented register. The exchanges were coded against behavioural codebooks, and a secondary coding audit checked the primary coding. This repository deposits the complete experiment histories, the codebooks, the section-by-section analysis data, and the audit report.
🗂️ Contents
| Folder | Contents |
|---|---|
transcripts/ |
The complete experiment histories for the three models (JSON). Typeset conversational-transcript PDFs of the same histories are in the OSF deposit. |
codebooks/ |
The per-model behavioural codebooks (JSON). PDF renderings are in the OSF deposit. |
data/ |
The analysis data tables by study section: classification results, features and correlations, model profiles, embeddings and cluster assignments, emergence and metacognitive-depth measures, covert user modelling, frame effectiveness, persona consistency, taboo patterns, resistance and acceptance, and the statistical results. Raw embedding arrays (.npy) are omitted in this text deposit; the embedding values are present in the CSV tables. |
results/ |
The final summary (JSON), the full analysis report (HTML), and the per-section metadata and summary files. |
coding-data/ |
The secondary coding audit report (plain-text rendering of the archived PDF). |
documentation/ |
The readme deposited with the OSF materials. |
Figure images from the OSF deposit are omitted in this text deposit; every plotted value is present in the tables.
📚 Citation
Please cite the article. Citation metadata is provided in CITATION.cff.
🔗 Access to the article
- Preprint (Research Square, open access): researchsquare.com/article/rs-10538113
- OSF deposit: osf.io/264qc
- Journal: submitted to AI and Society; the publisher link will be added here if the article is published.
⚖️ Licences
Data, transcripts, and documentation are released under the Creative Commons Attribution 4.0 International licence (CC BY 4.0, see LICENSE). Code is released under the MIT licence (see LICENSE-CODE).
The transcripts are outputs of publicly available AI systems, generated for this study. They contain no personal data.
- Downloads last month
- 104