Datasets:

wcamon
/

circus-0.3-eval-artifacts

Tags:

License:

Dataset card Data Studio Files Files and versions

xet

Community

Dataset Preview

Duplicate

Split (1)

train

The full dataset viewer is not available (click to read why). Only showing a preview of the rows.

The dataset generation failed

Error code:   DatasetGenerationError
Exception:    CastError
Message:      Couldn't cast
bfcl: struct<simple_ast: struct<broken: int64, fixed: int64, n: int64, post_accuracy: double, pre_accuracy (... 29 chars omitted)
  child 0, simple_ast: struct<broken: int64, fixed: int64, n: int64, post_accuracy: double, pre_accuracy: double, rows_file (... 9 chars omitted)
      child 0, broken: int64
      child 1, fixed: int64
      child 2, n: int64
      child 3, post_accuracy: double
      child 4, pre_accuracy: double
      child 5, rows_file: string
canaries: list<item: struct<bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int (... 4432 chars omitted)
  child 0, item: struct<bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_pe (... 4420 chars omitted)
      child 0, bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: doubl (... 4301 chars omitted)
          child 0, base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: doub (... 2049 chars omitted)
              child 0, rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: double, seqs: int (... 2022 chars omitted)
                  child 0, agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: double, seqs: int64, skip: int64, toke (... 10 chars omitted)
                      child 0, bpb: double
                      child 1, bytes: int64
                      child 2, nats_per_tok: double
  
...
    child 2, iter: int64
      child 3, kl: double
      child 4, learnable_frac: double
      child 5, loss_pg: double
      child 6, mean_reward: double
      child 7, n_groups: int64
      child 8, n_learnable: int64
      child 9, peak_gb: double
      child 10, ratio_clipfrac: double
      child 11, skipped: bool
      child 12, trunc_rate: double
schema_version: int64
sources: list<item: struct<path: string, sha256: string>>
  child 0, item: struct<path: string, sha256: string>
      child 0, path: string
      child 1, sha256: string
verdicts: list<item: struct<metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call:  (... 172 chars omitted)
  child 0, item: struct<metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call: double, gat (... 160 chars omitted)
      child 0, metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call: double, gate_c05: double, g (... 128 chars omitted)
          child 0, B_strict_band: double
          child 1, call_prior_delta: double
          child 2, copy_given_call: double
          child 3, gate_c05: double
          child 4, gate_c1: double
          child 5, gsm8k_pass8: double
          child 6, humaneval_pass8_obs: double
          child 7, repair_n: int64
          child 8, repair_rate: double
          child 9, humaneval_pass8: double
      child 1, stage: string
post_correct: bool
pre_correct: bool
post_error_class: string
id: string
pre_error_class: string
to
{'id': Value('string'), 'post_correct': Value('bool'), 'post_error_class': Value('string'), 'pre_correct': Value('bool'), 'pre_error_class': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1827, in _prepare_split_single
                  for key, table in generator:
                                    ^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
                  for item in generator(*args, **kwargs):
                              ~~~~~~~~~^^^^^^^^^^^^^^^^^
                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
              bfcl: struct<simple_ast: struct<broken: int64, fixed: int64, n: int64, post_accuracy: double, pre_accuracy (... 29 chars omitted)
                child 0, simple_ast: struct<broken: int64, fixed: int64, n: int64, post_accuracy: double, pre_accuracy: double, rows_file (... 9 chars omitted)
                    child 0, broken: int64
                    child 1, fixed: int64
                    child 2, n: int64
                    child 3, post_accuracy: double
                    child 4, pre_accuracy: double
                    child 5, rows_file: string
              canaries: list<item: struct<bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int (... 4432 chars omitted)
                child 0, item: struct<bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_pe (... 4420 chars omitted)
                    child 0, bpb: struct<base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: doubl (... 4301 chars omitted)
                        child 0, base: struct<rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: doub (... 2049 chars omitted)
                            child 0, rows: struct<agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: double, seqs: int (... 2022 chars omitted)
                                child 0, agentic_wake: struct<bpb: double, bytes: int64, nats_per_tok: double, secs: double, seqs: int64, skip: int64, toke (... 10 chars omitted)
                                    child 0, bpb: double
                                    child 1, bytes: int64
                                    child 2, nats_per_tok: double
                
              ...
                  child 2, iter: int64
                    child 3, kl: double
                    child 4, learnable_frac: double
                    child 5, loss_pg: double
                    child 6, mean_reward: double
                    child 7, n_groups: int64
                    child 8, n_learnable: int64
                    child 9, peak_gb: double
                    child 10, ratio_clipfrac: double
                    child 11, skipped: bool
                    child 12, trunc_rate: double
              schema_version: int64
              sources: list<item: struct<path: string, sha256: string>>
                child 0, item: struct<path: string, sha256: string>
                    child 0, path: string
                    child 1, sha256: string
              verdicts: list<item: struct<metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call:  (... 172 chars omitted)
                child 0, item: struct<metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call: double, gat (... 160 chars omitted)
                    child 0, metrics: struct<B_strict_band: double, call_prior_delta: double, copy_given_call: double, gate_c05: double, g (... 128 chars omitted)
                        child 0, B_strict_band: double
                        child 1, call_prior_delta: double
                        child 2, copy_given_call: double
                        child 3, gate_c05: double
                        child 4, gate_c1: double
                        child 5, gsm8k_pass8: double
                        child 6, humaneval_pass8_obs: double
                        child 7, repair_n: int64
                        child 8, repair_rate: double
                        child 9, humaneval_pass8: double
                    child 1, stage: string
              post_correct: bool
              pre_correct: bool
              post_error_class: string
              id: string
              pre_error_class: string
              to
              {'id': Value('string'), 'post_correct': Value('bool'), 'post_error_class': Value('string'), 'pre_correct': Value('bool'), 'pre_error_class': Value('string')}
              because column names don't match
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1880, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

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.

id string	post_correct bool	post_error_class string	pre_correct bool	pre_error_class string
simple_python_0	true	null	true	null
simple_python_1	true	null	false	ast_decoder:decoder_failed
simple_python_10	true	null	true	null
simple_python_100	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_101	false	simple_function_checker:missing_optional	false	simple_function_checker:unexpected_param
simple_python_102	false	simple_function_checker:missing_optional	false	value_error:string
simple_python_103	false	simple_function_checker:wrong_func_name	false	ast_decoder:decoder_failed
simple_python_104	true	null	true	null
simple_python_105	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_106	true	null	true	null
simple_python_107	true	null	true	null
simple_python_108	false	simple_function_checker:missing_optional	false	simple_function_checker:missing_optional
simple_python_109	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_11	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_110	false	simple_function_checker:wrong_func_name	true	null
simple_python_111	false	simple_function_checker:missing_required	false	simple_function_checker:missing_required
simple_python_112	false	value_error:others	true	null
simple_python_113	false	simple_function_checker:wrong_func_name	false	ast_decoder:decoder_failed
simple_python_114	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_115	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_116	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_117	false	value_error:others	false	simple_function_checker:wrong_func_name
simple_python_118	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_119	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_12	true	null	true	null
simple_python_120	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_121	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_122	true	null	false	ast_decoder:decoder_failed
simple_python_123	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_124	true	null	true	null
simple_python_125	true	null	true	null
simple_python_126	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_127	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_128	false	value_error:others	false	simple_function_checker:wrong_func_name
simple_python_129	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_13	false	type_error:nested	false	type_error:nested
simple_python_130	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_131	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_132	true	null	true	null
simple_python_133	false	ast_decoder:decoder_failed	false	value_error:others
simple_python_134	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_135	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_136	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_137	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_138	false	simple_function_checker:wrong_func_name	false	ast_decoder:decoder_failed
simple_python_139	false	simple_function_checker:wrong_func_name	false	ast_decoder:decoder_failed
simple_python_14	true	null	true	null
simple_python_140	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_141	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_142	true	null	true	null
simple_python_143	false	value_error:string	false	simple_function_checker:missing_required
simple_python_144	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_145	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_146	true	null	false	simple_function_checker:missing_required
simple_python_147	true	null	true	null
simple_python_148	false	value_error:others	false	simple_function_checker:wrong_func_name
simple_python_149	true	null	true	null
simple_python_15	false	simple_function_checker:wrong_func_name	false	ast_decoder:decoder_failed
simple_python_150	false	value_error:others	false	ast_decoder:decoder_failed
simple_python_151	false	value_error:others	false	simple_function_checker:missing_optional
simple_python_152	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_153	false	ast_decoder:decoder_failed	false	ast_decoder:decoder_failed
simple_python_154	false	ast_decoder:decoder_failed	false	simple_function_checker:missing_required
simple_python_155	false	value_error:others	false	ast_decoder:decoder_failed
simple_python_156	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_157	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_158	false	simple_function_checker:missing_required	false	simple_function_checker:missing_required
simple_python_159	false	value_error:string	false	simple_function_checker:missing_required
simple_python_16	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_160	true	null	false	simple_function_checker:wrong_func_name
simple_python_161	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_162	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_163	false	value_error:string	false	simple_function_checker:wrong_func_name
simple_python_164	false	simple_function_checker:missing_optional	true	null
simple_python_165	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_166	false	value_error:string	false	value_error:string
simple_python_167	true	null	false	simple_function_checker:wrong_func_name
simple_python_168	false	value_error:string	false	value_error:string
simple_python_169	true	null	false	simple_function_checker:wrong_func_name
simple_python_17	false	simple_function_checker:missing_required	true	null
simple_python_170	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_171	true	null	false	simple_function_checker:wrong_func_name
simple_python_172	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_173	false	value_error:string	false	value_error:string
simple_python_174	false	simple_function_checker:missing_required	false	simple_function_checker:wrong_func_name
simple_python_175	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_176	false	value_error:string	false	simple_function_checker:wrong_func_name
simple_python_177	true	null	true	null
simple_python_178	false	value_error:string	true	null
simple_python_179	true	null	false	simple_function_checker:wrong_func_name
simple_python_18	false	simple_function_checker:wrong_func_name	false	simple_function_checker:wrong_func_name
simple_python_180	true	null	true	null
simple_python_181	false	simple_function_checker:missing_optional	false	simple_function_checker:missing_optional
simple_python_182	true	null	true	null
simple_python_183	true	null	false	value_error:string
simple_python_184	true	null	false	simple_function_checker:wrong_func_name
simple_python_185	false	simple_function_checker:missing_optional	false	simple_function_checker:missing_optional
simple_python_186	false	value_error:string	false	simple_function_checker:wrong_func_name
simple_python_187	true	null	true	null
simple_python_188	true	null	false	simple_function_checker:wrong_func_name

End of preview.

circus-0.3 derived evaluation artifacts

This Dataset-type Hub repository contains the minimal, text-free evidence behind the paper “Elicit, Attribute, Repair.” It publishes aggregate metrics plus BFCL item IDs, paired correctness flags, coarse error classes, and hashes of the local source artifacts. It deliberately excludes prompts, model generations, reference answers, reasoning traces, trajectories, and raw training data.

What this is—and is not

This is a derived evaluation artifacts repository, not a new benchmark or a training dataset. It does not republish Hugging Face datasets, BFCL questions, function schemas, reference answers, or any upstream data rows. Hugging Face's “Dataset” repository type is used only because these reproducibility records are structured JSON/JSONL files.

Upstream benchmark

The item IDs refer to the official Berkeley Function Calling Leaderboard (BFCL) maintained in the Gorilla repository. Readers should obtain BFCL from that upstream project and follow the upstream benchmark's license and terms. The CC BY 4.0 metadata on this repository applies only to this release's original summaries, annotations, documentation, and derived correctness records; it does not relicense upstream benchmark content.

summary.json recomputes the headline BFCL Simple AST comparison. The bfcl_simple_paired.jsonl rows support the paired fixed/broken counts and an independent McNemar test. Five frozen-stage verdicts, five GSPO canary snapshots, the four-arm mechanism factorial, and numeric-only GSPO telemetry are included.

Headline result

The pre-repair checkpoint scores 20.25% and the post-repair checkpoint scores 31.25% over the same 400 BFCL Simple Python items: 56 items are fixed and 12 are broken.

Limitations

IDs require the separately obtained BFCL benchmark to interpret. The experiment is a single-seed, single-architecture case study and has no interactive evaluation.

Downloads last month: 49

Total file size:

158 kB

Collection including wcamon/circus-0.3-eval-artifacts

circus-0.3: Elicit, Attribute, Repair

Collection

Paper and minimized evidence for circus-0.3 agentic binding diagnosis and GSPO repair. Models remain gated. • 2 items • Updated 14 days ago