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 2 new columns ({'task_id', 'beta_hat'}) and 2 missing columns ({'exposure', 'onet_soc_code'}).
This happened while the csv dataset builder was generating data using
hf://datasets/MIT-WAL/evidence-grounded-ai-exposure/task_exposure_beta_hat.csv (at revision 331551e92c893e9217821a49437d9f13acfa09c3), ['hf://datasets/MIT-WAL/evidence-grounded-ai-exposure@331551e92c893e9217821a49437d9f13acfa09c3/occupation_exposure.csv', 'hf://datasets/MIT-WAL/evidence-grounded-ai-exposure@331551e92c893e9217821a49437d9f13acfa09c3/task_exposure_beta_hat.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 1837, in _prepare_split_single
writer.write_table(table)
~~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/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/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
pa_table = table_cast(pa_table, self._schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
task_id: int64
beta_hat: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 509
to
{'onet_soc_code': Value('string'), 'exposure': 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 1369, 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 948, 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 1683, 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 1839, 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 2 new columns ({'task_id', 'beta_hat'}) and 2 missing columns ({'exposure', 'onet_soc_code'}).
This happened while the csv dataset builder was generating data using
hf://datasets/MIT-WAL/evidence-grounded-ai-exposure/task_exposure_beta_hat.csv (at revision 331551e92c893e9217821a49437d9f13acfa09c3), ['hf://datasets/MIT-WAL/evidence-grounded-ai-exposure@331551e92c893e9217821a49437d9f13acfa09c3/occupation_exposure.csv', 'hf://datasets/MIT-WAL/evidence-grounded-ai-exposure@331551e92c893e9217821a49437d9f13acfa09c3/task_exposure_beta_hat.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.
onet_soc_code string | exposure float64 |
|---|---|
11-1011.00 | 0.309901 |
11-1011.03 | 0.495655 |
11-1021.00 | 0.482939 |
11-1031.00 | 0.376334 |
11-2011.00 | 0.296875 |
11-2021.00 | 0.644585 |
11-2022.00 | 0.293171 |
11-2032.00 | 0.581793 |
11-2033.00 | 0.521294 |
11-3012.00 | 0.407555 |
11-3013.00 | 0.220423 |
11-3013.01 | 0.329971 |
11-3021.00 | 0.637886 |
11-3031.00 | 0.346209 |
11-3031.01 | 0.449084 |
11-3031.03 | 0.774521 |
11-3051.00 | 0.451898 |
11-3051.01 | 0.408778 |
11-3051.02 | 0.241108 |
11-3051.03 | 0.139505 |
11-3051.04 | 0.196625 |
11-3051.06 | 0.29283 |
11-3061.00 | 0.589023 |
11-3071.00 | 0.269923 |
11-3071.04 | 0.491385 |
11-3111.00 | 0.46644 |
11-3121.00 | 0.356488 |
11-3131.00 | 0.535942 |
11-9013.00 | 0.171782 |
11-9021.00 | 0.154676 |
11-9031.00 | 0.223329 |
11-9032.00 | 0.180262 |
11-9033.00 | 0.339993 |
11-9041.00 | 0.374275 |
11-9041.01 | 0.335887 |
11-9051.00 | 0.20708 |
11-9071.00 | 0.325464 |
11-9072.00 | 0.178923 |
11-9081.00 | 0.171121 |
11-9111.00 | 0.353611 |
11-9121.00 | 0.365473 |
11-9121.01 | 0.387953 |
11-9121.02 | 0.576592 |
11-9131.00 | 0.24319 |
11-9141.00 | 0.228847 |
11-9151.00 | 0.227353 |
11-9161.00 | 0.389331 |
11-9171.00 | 0.191762 |
11-9179.01 | 0.290687 |
11-9179.02 | 0.410228 |
11-9199.01 | 0.53238 |
11-9199.02 | 0.487074 |
11-9199.08 | 0.350522 |
11-9199.09 | 0.231459 |
11-9199.10 | 0.45205 |
11-9199.11 | 0.520797 |
13-1011.00 | 0.485823 |
13-1021.00 | 0.441318 |
13-1022.00 | 0.374308 |
13-1023.00 | 0.671527 |
13-1031.00 | 0.43418 |
13-1032.00 | 0.209147 |
13-1041.00 | 0.466365 |
13-1041.01 | 0.368118 |
13-1041.03 | 0.515749 |
13-1041.04 | 0.284586 |
13-1041.06 | 0.15397 |
13-1041.07 | 0.683533 |
13-1041.08 | 0.505619 |
13-1051.00 | 0.531405 |
13-1071.00 | 0.694281 |
13-1074.00 | 0.187109 |
13-1075.00 | 0.438715 |
13-1081.00 | 0.425443 |
13-1081.01 | 0.635886 |
13-1081.02 | 0.611709 |
13-1082.00 | 0.586518 |
13-1111.00 | 0.597492 |
13-1121.00 | 0.362182 |
13-1131.00 | 0.617271 |
13-1141.00 | 0.550261 |
13-1151.00 | 0.584875 |
13-1161.00 | 0.887488 |
13-1161.01 | 0.764941 |
13-1199.04 | 0.707434 |
13-1199.05 | 0.573227 |
13-1199.06 | 0.65506 |
13-1199.07 | 0.396176 |
13-2011.00 | 0.56873 |
13-2022.00 | 0.434171 |
13-2023.00 | 0.625939 |
13-2031.00 | 0.730678 |
13-2041.00 | 0.628631 |
13-2051.00 | 0.600766 |
13-2052.00 | 0.529769 |
13-2053.00 | 0.572879 |
13-2054.00 | 0.676879 |
13-2061.00 | 0.437406 |
13-2071.00 | 0.516838 |
13-2072.00 | 0.479162 |
End of preview.
Evidence-Grounded Occupational AI Exposure
Calibrated AI exposure scores from the thesis Evidence-Grounded Measurement of Occupational Exposure to Artificial Intelligence (EPFL/MIT, 2026). Scores are produced by an ensemble of seven open-weight reasoning models judging ONET occupation–task pairs under a 2026 agentic-AI rubric, conditioned on retrieved news and research evidence plus ONET workplace descriptors, and calibrated against observed AI usage via simplex-constrained weighted least squares.
Vintage: 2026-07-20 pipeline release.
Files
task_exposure_beta_hat.csv — 18,205 rows
| column | description |
|---|---|
task_id |
O*NET 30.2 Task ID |
beta_hat |
Calibrated task-level exposure score in [0, 1] |
Task statements shared across occupations receive the same score. O*NET tasks outside the calibration sample (591 of 18,796) are omitted.
occupation_exposure.csv — 923 rows
| column | description |
|---|---|
onet_soc_code |
O*NET-SOC occupation code (e.g. 11-1011.00) |
exposure |
Occupation-level exposure in [0, 1]: task scores aggregated with O*NET task-time shares |
Interpretation
Scores estimate present-day task applicability: whether access to a contemporary AI system could reduce task completion time by at least half at equivalent quality (following Eloundou et al., 2024). They are time-stamped estimates tied to the evidence corpus and usage data of their vintage, not a forecast, and are re-estimated as evidence changes.
- Downloads last month
- 52