Datasets:

Aevion-Verifiable-AI
/

man-in-the-arena-leaderboard

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
chain_id: string
created_at: string
version: string
signing_method: string
public_key: string
entry_count: int64
last_updated: string
head_hash: string
signature: string
entry_id: string
sequence_number: int64
chain_hash: string
timestamp: string
proof_hash: string
proof: struct<proof_id: string, claim: string, verdict: string, confidence: double, variance_sigma: double, (... 202 chars omitted)
  child 0, proof_id: string
  child 1, claim: string
  child 2, verdict: string
  child 3, confidence: double
  child 4, variance_sigma: double
  child 5, risk_level: string
  child 6, evidence: list<item: string>
      child 0, item: string
  child 7, model_responses: list<item: struct<model: string, answer: string, confidence: double, agrees_with_consensus: bool>>
      child 0, item: struct<model: string, answer: string, confidence: double, agrees_with_consensus: bool>
          child 0, model: string
          child 1, answer: string
          child 2, confidence: double
          child 3, agrees_with_consensus: bool
  child 8, timestamp: string
  child 9, domain: string
previous_hash: string
to
{'entry_id': Value('string'), 'sequence_number': Value('int64'), 'timestamp': Value('string'), 'proof': {'proof_id': Value('string'), 'claim': Value('string'), 'verdict': Value('string'), 'confidence': Value('float64'), 'variance_sigma': Value('float64'), 'risk_level': Value('string'), 'evidence': List(Value('string')), 'model_responses': List({'model': Value('string'), 'answer': Value('string'), 'confidence': Value('float64'), 'agrees_with_consensus': Value('bool')}), 'timestamp': Value('string'), 'domain': Value('string')}, 'proof_hash': Value('string'), 'previous_hash': Value('string'), 'chain_hash': Value('string'), 'signature': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                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 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 295, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2281, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2227, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              chain_id: string
              created_at: string
              version: string
              signing_method: string
              public_key: string
              entry_count: int64
              last_updated: string
              head_hash: string
              signature: string
              entry_id: string
              sequence_number: int64
              chain_hash: string
              timestamp: string
              proof_hash: string
              proof: struct<proof_id: string, claim: string, verdict: string, confidence: double, variance_sigma: double, (... 202 chars omitted)
                child 0, proof_id: string
                child 1, claim: string
                child 2, verdict: string
                child 3, confidence: double
                child 4, variance_sigma: double
                child 5, risk_level: string
                child 6, evidence: list<item: string>
                    child 0, item: string
                child 7, model_responses: list<item: struct<model: string, answer: string, confidence: double, agrees_with_consensus: bool>>
                    child 0, item: struct<model: string, answer: string, confidence: double, agrees_with_consensus: bool>
                        child 0, model: string
                        child 1, answer: string
                        child 2, confidence: double
                        child 3, agrees_with_consensus: bool
                child 8, timestamp: string
                child 9, domain: string
              previous_hash: string
              to
              {'entry_id': Value('string'), 'sequence_number': Value('int64'), 'timestamp': Value('string'), 'proof': {'proof_id': Value('string'), 'claim': Value('string'), 'verdict': Value('string'), 'confidence': Value('float64'), 'variance_sigma': Value('float64'), 'risk_level': Value('string'), 'evidence': List(Value('string')), 'model_responses': List({'model': Value('string'), 'answer': Value('string'), 'confidence': Value('float64'), 'agrees_with_consensus': Value('bool')}), 'timestamp': Value('string'), 'domain': Value('string')}, 'proof_hash': Value('string'), 'previous_hash': Value('string'), 'chain_hash': Value('string'), 'signature': Value('string')}
              because column names don't match

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.

Aevion Arena Leaderboard

This dataset is a cryptographically signed arena artifact bundle generated by the Aevion man_in_the_arena runner. It publishes the benchmark leaderboard, signed receipt, Wharf Performance Oracle manifest, and evidence-chain metadata as a replayable proof bundle instead of a plain score table.

Verified Artifact State

Leaderboard ID: arena-93448eaa4fa1
Receipt ID: 05028766b3afe7373063cd99ef2d9395
Attestation mode: arena-kernel
Top model: qwen3:8b
Merkle root: b0ea5d02263025283243fe2d876b35bc3c6b951a424a0025b64d4dc864664600
Chain verification: intact (2/2)
PQ algorithm: ML-DSA-65
Signature verified in receipt: False
Signature replay-verified now: False
Signing mode: ephemeral
Manifest signing status: signed_locally
Beacon anchor: pending:nist-anchor

Scoreboard

Rank	Model	Composite	Spectral Event Rate	Mean Latency (ms)	Mean TPS
1	qwen3:8b	0.833333	0.500000	7687.440	57.069

Evidence Chain

Chain ID: arena-20260422T123934Z
Head hash: unknown
Entry count: 0
Chain signing method: ed25519

Caveats

signature_verified records the result captured during the original run.
Signature replay-verified now reflects the current verifier replaying the receipt bytes in this runtime.
manifest_signing_status reflects whether the Wharf manifest is locally signed or waiting on Sheriff-backed enrollment.
beacon_anchor remains pending until hardware-rooted anchoring is enrolled.

Bundle Contents

leaderboard.json: structured benchmark ranking and chain summary
receipt.json: signed receipt payload with PQ metadata
leaderboard.xgml: graph representation of the proof bundle
performance_oracle_part.json: Wharf-ready manifest for the arena result
evidence_chain/*.jsonl: append-only case evidence log
evidence_chain/*.meta.json: chain hash summary

Reproduction

Run the arena locally and emit a new signed artifact bundle:

python3 -m core.python.verification.man_in_the_arena \
  --suite claim_verification_f1 \
  --model qwen3:8b \
  --output-dir .aevion/receipts/arena \
  --generate-readme

Publish an existing bundle to Hugging Face:

python3 -m core.python.verification.man_in_the_arena \
  --output-dir .aevion/receipts/arena \
  --publish-existing \
  --publish-leaderboard \
  --hf-repo aevionai/man-in-the-arena-leaderboard

Signature Material

Public key present: True
Signature value present: True

The goal of this dataset is evidence portability: the bundle can be inspected and replayed independently of the original runtime.

Downloads last month: 76

Total file size:

234 kB