The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
config: string
problem_id: string
difficulty: string
area: string
verified: bool
claimed_correct: bool
trace: string
lean_code: string
llm_calls: int64
tokens_in: int64
tokens_out: int64
latency_s: double
verification_is_posthoc: bool
model: string
to
{'config': Value('string'), 'problem_id': Value('string'), 'difficulty': Value('string'), 'area': Value('string'), 'verified': Value('bool'), 'claimed_correct': Value('bool'), 'trace': Value('string'), 'lean_code': Value('string'), 'llm_calls': Value('int64'), 'tokens_in': Value('int64'), 'tokens_out': Value('int64'), 'latency_s': Value('float64'), 'model': Value('string')}
because column names don't match
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 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
config: string
problem_id: string
difficulty: string
area: string
verified: bool
claimed_correct: bool
trace: string
lean_code: string
llm_calls: int64
tokens_in: int64
tokens_out: int64
latency_s: double
verification_is_posthoc: bool
model: string
to
{'config': Value('string'), 'problem_id': Value('string'), 'difficulty': Value('string'), 'area': Value('string'), 'verified': Value('bool'), 'claimed_correct': Value('bool'), 'trace': Value('string'), 'lean_code': Value('string'), 'llm_calls': Value('int64'), 'tokens_in': Value('int64'), 'tokens_out': Value('int64'), 'latency_s': Value('float64'), 'model': Value('string')}
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.
Dataset card
Contents
| File | Description |
|---|---|
data/eval_set.jsonl |
30 miniF2F problems selected for evaluation, with split and labels |
results/*.jsonl |
one record per problem and configuration (5 configurations × dev/test) |
Evaluation set
30 formal statements taken from miniF2F (Lean 4 version, statements only: no reference proofs are included). Families: amc12 (11), mathd (10), aime (5), imo (4).
| Basic | Intermediate | Advanced | Total | |
|---|---|---|---|---|
| Development | 6 | 2 | 2 | 10 |
| Test | 4 | 9 | 7 | 20 |
Areas (manual labels): algebra 11, counting 10, number theory 9. Development: algebra 5 / counting 2 / number theory 3; test: 6 / 7 / 7. The split is fixed: development problems were used to build and tune the system, test problems were evaluated once after the code was frozen.
Fields of eval_set.jsonl
id, source_family, difficulty (basico | intermedio | avanzado), area, area_source (nombre = from the
problem name, heuristica = keywords; the latter are less reliable), statement_nl (natural language),
statement_lean, lean_with_sorry, has_reference_proof, maintainer_note, prelude
(open scoped Nat Real), split (dev | test).
Results files
File names: baseline_a_<split>.jsonl, baseline_b_<split>.jsonl, pipeline_<config>_<split>_v1[_lean].jsonl
(_lean = RAG query built from the Lean statement). Main fields: problem_id, verified (accepted by Lean),
claimed_correct (delivered as correct), lean_code, trace (last Lean message), llm_calls, tokens_in,
tokens_out, latency_s; pipeline runs add subgoals, retrieved, r (replans), k, status, failure_reason.
Baseline B's verification is post-hoc (it never uses Lean to decide what to deliver).
Known issues and limitations
- Labels for area and difficulty are manual; some counting labels are debatable (see
data/auditoria.md). - Formal statements are used as in miniF2F; Lean checks the proof against the formal statement, not against the natural-language problem. Mis-formalizations are possible.
- miniF2F is public and its proofs circulate online: training-data contamination of the LLM cannot be ruled out.
- 30 problems and one run per configuration give very wide confidence intervals.
- Results depend on Lean 4.36.0-rc2, Mathlib commit
9d581dfff2e41de989072ea63abf3daf69311520andgemini-3.5-flash-lite(temperature 0). Error categories are heuristic.
Licenses and attribution
miniF2F (Zheng, Han, Polu, 2022) and Mathlib have their own licenses [verificar y citar]. Generated proofs and analysis: [licencia por definir].
Citation
Figueroa Zapata, K. N. and Páez Alarcón, A. Multi-agent generative AI with formal verification in Lean 4. UPTC Sogamoso, 2026. [completar con la referencia final del informe]
- Downloads last month
- 21