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
seed: int64
total: double
steps: int64
reason: string
actions: list<item: int64>
child 0, item: int64
final: struct<step: int64, total: double, dead: bool, player: list<item: int64>, entities: list<item: list< (... 265 chars omitted)
child 0, step: int64
child 1, total: double
child 2, dead: bool
child 3, player: list<item: int64>
child 0, item: int64
child 4, entities: list<item: list<item: string>>
child 0, item: list<item: string>
child 0, item: string
child 5, last_action: int64
child 6, timers: struct<spawn_timer: int64, move_timer: int64, ramp_timer: int64, ramp_index: int64, spawn_speed: int (... 22 chars omitted)
child 0, spawn_timer: int64
child 1, move_timer: int64
child 2, ramp_timer: int64
child 3, ramp_index: int64
child 4, spawn_speed: int64
child 5, move_speed: int64
child 7, rng: struct<kind: string, keys: list<item: int64>, pos: int64, has_gauss: int64, cached: double>
child 0, kind: string
child 1, keys: list<item: int64>
child 0, item: int64
child 2, pos: int64
child 3, has_gauss: int64
child 4, cached: double
target_corrected: int64
reasons: struct<target_not_corrected: int64, regression_gate: int64>
child 0, target_not_corrected: int64
child 1, regression_gate: int64
both_means_improved: int64
revision: string
summary: struct<status: string, accepted_checkpoints: int64, proposals: int64, cycles: int64, base_hash: stri (... 536 chars omitted)
child 0, status: string
child 1, accepted_checkpoints: int64
child 2, proposals: int64
child 3, cycles: int64
child 4, base_hash: string
child 5, final_hash: string
child 6, development: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 7, gate: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 8, held_out_audit: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 9, final_replay_passed: bool
child 10, retained_repairs: list<item: null>
child 0, item: null
child 11, scope: string
child 12, publishable: bool
to
{'revision': Value('string'), 'reasons': {'target_not_corrected': Value('int64'), 'regression_gate': Value('int64')}, 'both_means_improved': Value('int64'), 'target_corrected': Value('int64'), 'summary': {'status': Value('string'), 'accepted_checkpoints': Value('int64'), 'proposals': Value('int64'), 'cycles': Value('int64'), 'base_hash': Value('string'), 'final_hash': Value('string'), 'development': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'gate': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'held_out_audit': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'final_replay_passed': Value('bool'), 'retained_repairs': List(Value('null')), 'scope': Value('string'), 'publishable': Value('bool')}}
because column names don't match
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 294, in _generate_tables
pa_table.validate()
~~~~~~~~~~~~~~~~~^^
File "pyarrow/table.pxi", line 4237, in pyarrow.lib.Table.validate
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
pyarrow.lib.ArrowInvalid: Column 5: In chunk 0: Invalid: Struct child array #4 invalid: Invalid: Length spanned by list offsets (280) larger than values array (length 279)
During handling of the above exception, another exception occurred:
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 332, 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 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
seed: int64
total: double
steps: int64
reason: string
actions: list<item: int64>
child 0, item: int64
final: struct<step: int64, total: double, dead: bool, player: list<item: int64>, entities: list<item: list< (... 265 chars omitted)
child 0, step: int64
child 1, total: double
child 2, dead: bool
child 3, player: list<item: int64>
child 0, item: int64
child 4, entities: list<item: list<item: string>>
child 0, item: list<item: string>
child 0, item: string
child 5, last_action: int64
child 6, timers: struct<spawn_timer: int64, move_timer: int64, ramp_timer: int64, ramp_index: int64, spawn_speed: int (... 22 chars omitted)
child 0, spawn_timer: int64
child 1, move_timer: int64
child 2, ramp_timer: int64
child 3, ramp_index: int64
child 4, spawn_speed: int64
child 5, move_speed: int64
child 7, rng: struct<kind: string, keys: list<item: int64>, pos: int64, has_gauss: int64, cached: double>
child 0, kind: string
child 1, keys: list<item: int64>
child 0, item: int64
child 2, pos: int64
child 3, has_gauss: int64
child 4, cached: double
target_corrected: int64
reasons: struct<target_not_corrected: int64, regression_gate: int64>
child 0, target_not_corrected: int64
child 1, regression_gate: int64
both_means_improved: int64
revision: string
summary: struct<status: string, accepted_checkpoints: int64, proposals: int64, cycles: int64, base_hash: stri (... 536 chars omitted)
child 0, status: string
child 1, accepted_checkpoints: int64
child 2, proposals: int64
child 3, cycles: int64
child 4, base_hash: string
child 5, final_hash: string
child 6, development: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 7, gate: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 8, held_out_audit: struct<before_mean: double, after_mean: double, gain: double, wins: int64, losses: int64, ties: int6 (... 25 chars omitted)
child 0, before_mean: double
child 1, after_mean: double
child 2, gain: double
child 3, wins: int64
child 4, losses: int64
child 5, ties: int64
child 6, lost_survivors: int64
child 9, final_replay_passed: bool
child 10, retained_repairs: list<item: null>
child 0, item: null
child 11, scope: string
child 12, publishable: bool
to
{'revision': Value('string'), 'reasons': {'target_not_corrected': Value('int64'), 'regression_gate': Value('int64')}, 'both_means_improved': Value('int64'), 'target_corrected': Value('int64'), 'summary': {'status': Value('string'), 'accepted_checkpoints': Value('int64'), 'proposals': Value('int64'), 'cycles': Value('int64'), 'base_hash': Value('string'), 'final_hash': Value('string'), 'development': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'gate': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'held_out_audit': {'before_mean': Value('float64'), 'after_mean': Value('float64'), 'gain': Value('float64'), 'wins': Value('int64'), 'losses': Value('int64'), 'ties': Value('int64'), 'lost_survivors': Value('int64')}, 'final_replay_passed': Value('bool'), 'retained_repairs': List(Value('null')), 'scope': Value('string'), 'publishable': Value('bool')}}
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.
Asterix recursive learning experiment
New game and checkpoint; transfer of method, not transfer of Breakout weights. Native MinAtar sticky actions 0.1 retained; policy greedy; 1000-step pilot differs from upstream sampled 10000-step evaluation.
Run protocol: frozen source.
Development and gate seeds are reused for selection. The final audit is evaluated after selection ends. All candidates and failed attempts are retained. Base weights and upstream material: Rikiya Takehi and collaborators, CC BY 4.0. Official MinAtar game: Kenny Young and Tian Tian, GPL-3.0.
- Downloads last month
- 139