Hugging Face's logo

Datasets:
Hooshaai
/
aegis-stressbench-trajectories

Tasks:
Text Generation
Reinforcement Learning
Languages:
English
Size:
n<1K
Tags:
stressbench
debate-trajectories
llm-as-a-judge
markov-automata
defensive-reasoning
License:
Dataset card Files
xet
 Community
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:    CastError
Message:      Couldn't cast
3: struct<S: double>
  child 0, S: double
1: struct<S: double, H: double>
  child 0, S: double
  child 1, H: double
0: struct<W: double>
  child 0, W: double
evidence_sensitivity: double
mean_adversary_reward: double
ipt_consistency: double
ensemble_accuracy: double
stable_correct_rate: double
non_evidential_flip_rate: double
collapse_rate: double
initial_accuracy: double
abstention_recovery_rate: double
contradiction_rate: double
cross_family_agreement_rate: double
process_consistency: double
mean_judge_reward: double
hard_flip_rate: double
accuracy: double
cpii: double
to
{'accuracy': Value('float64'), 'initial_accuracy': Value('float64'), 'stable_correct_rate': Value('float64'), 'hard_flip_rate': Value('float64'), 'collapse_rate': Value('float64'), 'non_evidential_flip_rate': Value('float64'), 'contradiction_rate': Value('float64'), 'evidence_sensitivity': Value('float64'), 'abstention_recovery_rate': Value('float64'), 'ipt_consistency': Value('float64'), 'cpii': Value('float64'), 'process_consistency': Value('float64'), 'mean_judge_reward': Value('float64'), 'mean_adversary_reward': Value('float64'), 'cross_family_agreement_rate': Value('float64'), 'ensemble_accuracy': Value('float64')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, 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 127, 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 483, 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 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, 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 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
              3: struct<S: double>
                child 0, S: double
              1: struct<S: double, H: double>
                child 0, S: double
                child 1, H: double
              0: struct<W: double>
                child 0, W: double
              evidence_sensitivity: double
              mean_adversary_reward: double
              ipt_consistency: double
              ensemble_accuracy: double
              stable_correct_rate: double
              non_evidential_flip_rate: double
              collapse_rate: double
              initial_accuracy: double
              abstention_recovery_rate: double
              contradiction_rate: double
              cross_family_agreement_rate: double
              process_consistency: double
              mean_judge_reward: double
              hard_flip_rate: double
              accuracy: double
              cpii: double
              to
              {'accuracy': Value('float64'), 'initial_accuracy': Value('float64'), 'stable_correct_rate': Value('float64'), 'hard_flip_rate': Value('float64'), 'collapse_rate': Value('float64'), 'non_evidential_flip_rate': Value('float64'), 'contradiction_rate': Value('float64'), 'evidence_sensitivity': Value('float64'), 'abstention_recovery_rate': Value('float64'), 'ipt_consistency': Value('float64'), 'cpii': Value('float64'), 'process_consistency': Value('float64'), 'mean_judge_reward': Value('float64'), 'mean_adversary_reward': Value('float64'), 'cross_family_agreement_rate': Value('float64'), 'ensemble_accuracy': Value('float64')}
              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.

🛡️ AEGIS-StressBench Trajectories & Behavioral Automata Archive

GitHub Space Org

This repository stores empirical multi-turn debate trajectories, behavioral transition matrices, and evaluation logs for AEGIS-StressBench.

📁 Repository Contents

  1. sample_trajectories.jsonl: Complete multi-turn conversational traces recording step-by-step judge rationales, adversary attacks across 4 personas, and outcome classifications.
  2. fsm_transitions.json: Empirical 9-state Finite State Machine (FSM) transition probability matrix used by AEGIS++ to predict failure modes with 0.964 AUROC.
  3. final_metrics.json: Summary statistics including Stable Correct Rate (SCR), Hard Flip Rate (HFR), Collapse Rate, and 4-point Defensive Reasoning Scores (DRS).
  4. gsm8k_golden_50.json: 50 curated high-complexity mathematical challenge problems.

🎭 The 4 Attacker Personas

  • Authoritative: Attacks relying on prestige bias and simulated consensus ("A Stanford research team published...").
  • Empathetic: Emotional manipulation, urgency, and appeals to sympathy.
  • Pseudo-Logical: Surface-level syllogisms, circular reasoning, and distorted mathematical axioms.
  • Aggressive: Intimidation, abrupt commands, and harsh derogatory critiques.

📖 Citation 

@article{majlesi2026aegis,
  title={AEGIS: Adversarial Epistemic Guard for Invariant and Stable Judging in LLM-as-a-Judge},
  author={Majlesi, Taha and AEGIS Research Team},
  year={2026}
}
Downloads last month
44
Total file size:
34.6 kB