Datasets:

publicai-dev
/

Trajector-SWE-20K-sample

Error code:   StreamingRowsError
Exception:    TypeError
Message:      Couldn't cast array of type
struct<description: string, input_schema: string, name: string, cache_control: struct<type: string>>
to
{'description': Value('string'), 'input_schema': Json(decode=True), 'name': Value('string')}
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_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 2315, in cast_table_to_schema
                  cast_array_to_feature(
                  ~~~~~~~~~~~~~~~~~~~~~^
                      table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                      feature,
                      ^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1861, in wrapper
                  return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
                                           ~~~~^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2121, in cast_array_to_feature
                  casted_array_values = _c(array.values, feature.feature)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1863, in wrapper
                  return func(array, *args, **kwargs)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2161, in cast_array_to_feature
                  raise TypeError(f"Couldn't cast array of type\n{_short_str(array.type)}\nto\n{_short_str(feature)}")
              TypeError: Couldn't cast array of type
              struct<description: string, input_schema: string, name: string, cache_control: struct<type: string>>
              to
              {'description': Value('string'), 'input_schema': Json(decode=True), 'name': Value('string')}

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.

Trajector-SWE-20K — open sample (2,000 sessions)

Collected with the Trajector plugin for coding agents — https://trajector.publicai.io · Paper: Actions Without Reasons (NeurIPS 2026) · Code: https://github.com/PublicAI01/action-grounded-reasoning · Full release (20,000 sessions; the remaining 18,000 are available on request): Trajector-SWE-20K

This repository contains 2,000 of the 20,000 sessions of Trajector-SWE-20K and is openly downloadable. The other 18,000 sessions are in the main repository, where the dataset card, the full statistics, the consent terms and the redaction description live. The two parts do not overlap.

How the sample was drawn

Sessions were taken from the 20,000-session release in hash order (sha256("sample2k|" + session_id)). Each candidate then went through an additional line-by-line screen: every line containing a credential keyword was sent to a language model with random-looking values masked, flagged lines were triaged by an AI assistant on the masked text without human review, and sessions judged a plausible residual were left out. The screen does not see credentials or personal data on lines without a keyword. The sample is therefore biased towards sessions with fewer credential-like strings; see CHECKS.md.

At a glance (this sample)


Sessions	2,000
Assistant turns	78,053
Tool calls	78,273
Estimated tokens	≈89M (chars/4 + CJK chars/1.5 on the released text)
Non-empty reasoning per 100 assistant turns	17.8
Sessions with no non-empty reasoning	710 (35.5%)

Statistics of the whole release are in STATS_release_20000.json and on the main card.

Data format

Four gzip-compressed JSON Lines shards under data/, one session per line, with the fields session_id, model, thinking_effort, system, tools, messages (see the main card for details). MANIFEST.txt lists size and SHA-256 of each shard.

Consent, redaction, use

Identical to the main release: sessions come from developers who accepted an authorization naming public release among the permitted uses; credentials, contact details and embedded files are replaced by a frozen pattern-based pipeline and flagged sessions are removed as a whole. Redaction is pattern-based and not exhaustive; if you find a residual credential or personal detail, write to opensource@publicai.io and the session is removed in the next revision. Do not use the data to identify individuals.

License: ODC-BY. Please do not re-host or re-upload this sample elsewhere: link to this repository instead, so that revisions (for example after a withdrawal request) reach every user. The full release is available on request from the main repository and is not to be redistributed.

Citation

@inproceedings{wang2026actions,
  title     = {Actions Without Reasons: Filling the Reasoning Gap in a Large-Scale Public Dataset of Real Coding-Agent Trajectories},
  author    = {Wang, Qin and Owji, Afsoun and Yin, Hang and Liu, Guangyi},
  booktitle = {NeurIPS 2026 Workshop: Transitioning from Pre-Training to Post-Training},
  year      = {2026}
}

Downloads last month: 105

Total file size:

212 MB