Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
greeting_1: string
greeting_2: string
messages_1: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
messages_2: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
conversation: list<item: string>
child 0, item: string
messages: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
model_key: string
stage: string
files: struct<self_interaction-leading.jsonl: string, self_interaction.jsonl: string, self_reflection.jsonl (... 33 chars omitted)
child 0, self_interaction-leading.jsonl: string
child 1, self_interaction.jsonl: string
child 2, self_reflection.jsonl: string
child 3, sft_data.jsonl: string
n_interact: int64
fork_sha: string
n_reflect: int64
turns: int64
seed: int64
dpo_adapter_sha256_12: string
to
{'dpo_adapter_sha256_12': Value('string'), 'files': {'self_interaction-leading.jsonl': Value('string'), 'self_interaction.jsonl': Value('string'), 'self_reflection.jsonl': Value('string'), 'sft_data.jsonl': Value('string')}, 'fork_sha': Value('string'), 'model_key': Value('string'), 'n_interact': Value('int64'), 'n_reflect': Value('int64'), 'seed': Value('int64'), 'stage': Value('string'), 'turns': Value('int64')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, 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 127, 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 483, 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 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, 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 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, 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 343, 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 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
greeting_1: string
greeting_2: string
messages_1: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
messages_2: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
conversation: list<item: string>
child 0, item: string
messages: list<item: struct<role: string, content: string>>
child 0, item: struct<role: string, content: string>
child 0, role: string
child 1, content: string
model_key: string
stage: string
files: struct<self_interaction-leading.jsonl: string, self_interaction.jsonl: string, self_reflection.jsonl (... 33 chars omitted)
child 0, self_interaction-leading.jsonl: string
child 1, self_interaction.jsonl: string
child 2, self_reflection.jsonl: string
child 3, sft_data.jsonl: string
n_interact: int64
fork_sha: string
n_reflect: int64
turns: int64
seed: int64
dpo_adapter_sha256_12: string
to
{'dpo_adapter_sha256_12': Value('string'), 'files': {'self_interaction-leading.jsonl': Value('string'), 'self_interaction.jsonl': Value('string'), 'self_reflection.jsonl': Value('string'), 'sft_data.jsonl': Value('string')}, 'fork_sha': Value('string'), 'model_key': Value('string'), 'n_interact': Value('int64'), 'n_reflect': Value('int64'), 'seed': Value('int64'), 'stage': Value('string'), 'turns': Value('int64')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Status: SUPERSEDED by
Misalignment-Empirics/theo_organism-training-dataas the place for training data (2026-10-01, canonical copies of thekeep/impulsive-glmv3/introspection data moved there). Still the OCT runbook's working cache; kept as provenance.
theo_oct-behaviour-data — OCT runbook cache (not a training starting point)
Default $HF_DATA of scripts/runbook_oct.sh: the runbook publishes and re-fetches its
per-run intermediate data here, for any persona and base model. Layout:
keep/<persona>[-<run tag>]/{teacher,student,introspection}/<model key>/….
keep/impulsive-glmv3/introspection/qwen-2.5-{7b,14b,32b}-it/<dpo-sha>/— self-reflection, self-interaction and introspection-SFT data of the registered v3 impulsiveoct_behaviourorganisms (evaluated 2026-09-25). Keyed by the DPO adapter's sha: generated BY that adapter, so it is provenance, not reusable for another base model or behaviour.- (
keep/impulsive/, the retired gpt-4o-teacher pre-v3 build, was deleted 2026-09-29; no registered organism used it.)
To TRAIN impulsive organisms on a new base model, start from
Misalignment-Empirics/theo_oct-glm-v3-training-data (see its README).
- Downloads last month
- 206