Dataset Viewer
Duplicate
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:    TypeError
Message:      Couldn't cast array of type string to null
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 489, 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 2818, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, 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 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2303, 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 1852, 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 2109, 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 1854, in wrapper
                  return func(array, *args, **kwargs)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2143, in cast_array_to_feature
                  return array_cast(
                      array,
                  ...<2 lines>...
                      allow_decimal_to_str=allow_decimal_to_str,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1854, in wrapper
                  return func(array, *args, **kwargs)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2005, in array_cast
                  raise TypeError(f"Couldn't cast array of type {_short_str(array.type)} to {_short_str(pa_type)}")
              TypeError: Couldn't cast array of type string to null

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.

GSPC Boards — 13-axis durable board

Deterministic multi-model boards for the 13 GSPC axes, produced by sovos-city's durable_board runner. Each board is pushed here the instant its axis finishes, so no pod reboot can lose a completed axis (this repo survived several this session; the pods did not). Per-item rows are published alongside each board so every number is recomputable.

Honest status (2026-08-12)

Run fleet: 5 surviving sovereign models on a stable 3090 (sov-safety-v1, sov-merge-slerp, sov-merge-dare, sov-refusal-combo-lora, qwen2.5 base). The exact sov6-v3-light fleet was ephemeral and lost in a pod reboot — these are the models that landed somewhere durable.

Axis Status Note
governance MEASURED (n=237) sovereign best 0.392 > base 0.291
safety, provenance, continuity, conformance, openness, machinery, care, cross-reality, detector-interop, art5-safeguard, swarm, affect UNMEASURED the small sovereign fleet could not produce quotable label responses — reported honestly, not scored

Only governance measured. The other 12 are UNMEASURED because a weak fleet could not answer in the label schema — the board says so rather than inventing a number. A richer instruction-tuned fleet (gemma3:12b etc.) re-measures the UNMEASURED axes on resume; durable_board skips only MEASURED boards.

Tail risk — governance (the fat tail the mean hides)

Computed by sovos-city.tail@1.1.0 over the published per-item rows (reproducible by anyone):

statistic value meaning
mean item-pass 0.297 the linear number — looks ~30% competent
CVaR-5% 0.000 the worst 5% of items: zero models pass
correlated-failure 48.3% ~half the items break the whole fleet at once

The mean says 30%; the tail says half the items are catastrophic and correlated — every model fails them together. Independent errors average out; correlated ones are the fat tail that executes at scale. A mean-only board would have hidden this entirely. Tail stats are quoted only at n>=100 (governance clears it); worst-case and correlated-failure are any-n signals.

Method

Deterministic exact-label grading; unparsed counts incorrect; transport failures excluded as ours; canaries excluded; nothing quoted below usable n>=30. No model judges another model. Each board names its aggregator (name@version(params)) so the same rows cannot be quietly re-aggregated into a different number.

Downloads last month
-