Datasets:

sailab-vienna
/

privesc-llm-data

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
messages: list<item: struct<role: string, content: string, tool_calls: list<item: struct<type: string, id: str (... 94 chars omitted)
  child 0, item: struct<role: string, content: string, tool_calls: list<item: struct<type: string, id: string, functi (... 82 chars omitted)
      child 0, role: string
      child 1, content: string
      child 2, tool_calls: list<item: struct<type: string, id: string, function: struct<name: string, arguments: string>>>
          child 0, item: struct<type: string, id: string, function: struct<name: string, arguments: string>>
              child 0, type: string
              child 1, id: string
              child 2, function: struct<name: string, arguments: string>
                  child 0, name: string
                  child 1, arguments: string
      child 3, tool_call_id: string
      child 4, name: string
scenario: string
num_tokens: int64
run_id: int64
success: bool
turns: int64
final_message: string
model: string
mode: string
cost: double
prompt_tokens: int64
completion_tokens: int64
tools: list<item: struct<type: string, function: struct<name: string, description: string, parameters: stru (... 223 chars omitted)
  child 0, item: struct<type: string, function: struct<name: string, description: string, parameters: struct<properti (... 211 chars omitted)
      child 0, type: string
      child 1, function: struct<name: string, description: string, parameters: struct<properties: struct<command: struct<desc (... 179 chars omitted)
      
...
, capabilities_gtfobins: int64
num_runs: int64
traces_gt_64k: int64
scenario_avg_turns: struct<capabilities_gtfobins: double>
  child 0, capabilities_gtfobins: double
avg_tokens_per_run: double
traces_gt_16k: int64
scenario_long_ctx: struct<capabilities_gtfobins: struct<16k: int64, 32k: int64, 64k: int64>>
  child 0, capabilities_gtfobins: struct<16k: int64, 32k: int64, 64k: int64>
      child 0, 16k: int64
      child 1, 32k: int64
      child 2, 64k: int64
min_turns: int64
total_messages: int64
avg_toolcalls_per_run: double
max_tokens: int64
max_turns: int64
metadata_histograms: struct<generator_name: struct<capabilities_gtfobins: int64>, category: struct<capabilities_gtfobins: (... 153 chars omitted)
  child 0, generator_name: struct<capabilities_gtfobins: int64>
      child 0, capabilities_gtfobins: int64
  child 1, category: struct<capabilities_gtfobins: int64>
      child 0, capabilities_gtfobins: int64
  child 2, binary_name: struct<perl: int64, php: int64, ruby: int64>
      child 0, perl: int64
      child 1, php: int64
      child 2, ruby: int64
  child 3, binary_path: struct</usr/bin/perl: int64, /usr/bin/php: int64, /usr/bin/ruby: int64>
      child 0, /usr/bin/perl: int64
      child 1, /usr/bin/php: int64
      child 2, /usr/bin/ruby: int64
toolcall_types: struct<exec_command: int64, test_credentials: int64>
  child 0, exec_command: int64
  child 1, test_credentials: int64
avg_messages_per_run: double
traces_gt_32k: int64
min_tokens: int64
total_toolcalls: int64
to
{'num_runs': Value('int64'), 'total_messages': Value('int64'), 'avg_messages_per_run': Value('float64'), 'total_toolcalls': Value('int64'), 'avg_toolcalls_per_run': Value('float64'), 'total_tokens': Value('int64'), 'avg_tokens_per_run': Value('float64'), 'scenarios': {'capabilities_gtfobins': Value('int64')}, 'toolcall_types': {'exec_command': Value('int64'), 'test_credentials': Value('int64')}, 'traces_gt_16k': Value('int64'), 'traces_gt_32k': Value('int64'), 'traces_gt_64k': Value('int64'), 'scenario_long_ctx': {'capabilities_gtfobins': {'16k': Value('int64'), '32k': Value('int64'), '64k': Value('int64')}}, 'scenario_avg_turns': {'capabilities_gtfobins': Value('float64')}, 'min_turns': Value('int64'), 'max_turns': Value('int64'), 'avg_turns': Value('float64'), 'min_tokens': Value('int64'), 'max_tokens': Value('int64'), 'metadata_histograms': {'generator_name': {'capabilities_gtfobins': Value('int64')}, 'category': {'capabilities_gtfobins': Value('int64')}, 'binary_name': {'perl': Value('int64'), 'php': Value('int64'), 'ruby': Value('int64')}, 'binary_path': {'/usr/bin/perl': Value('int64'), '/usr/bin/php': Value('int64'), '/usr/bin/ruby': Value('int64')}}}
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 478, 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 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              messages: list<item: struct<role: string, content: string, tool_calls: list<item: struct<type: string, id: str (... 94 chars omitted)
                child 0, item: struct<role: string, content: string, tool_calls: list<item: struct<type: string, id: string, functi (... 82 chars omitted)
                    child 0, role: string
                    child 1, content: string
                    child 2, tool_calls: list<item: struct<type: string, id: string, function: struct<name: string, arguments: string>>>
                        child 0, item: struct<type: string, id: string, function: struct<name: string, arguments: string>>
                            child 0, type: string
                            child 1, id: string
                            child 2, function: struct<name: string, arguments: string>
                                child 0, name: string
                                child 1, arguments: string
                    child 3, tool_call_id: string
                    child 4, name: string
              scenario: string
              num_tokens: int64
              run_id: int64
              success: bool
              turns: int64
              final_message: string
              model: string
              mode: string
              cost: double
              prompt_tokens: int64
              completion_tokens: int64
              tools: list<item: struct<type: string, function: struct<name: string, description: string, parameters: stru (... 223 chars omitted)
                child 0, item: struct<type: string, function: struct<name: string, description: string, parameters: struct<properti (... 211 chars omitted)
                    child 0, type: string
                    child 1, function: struct<name: string, description: string, parameters: struct<properties: struct<command: struct<desc (... 179 chars omitted)
                    
              ...
              , capabilities_gtfobins: int64
              num_runs: int64
              traces_gt_64k: int64
              scenario_avg_turns: struct<capabilities_gtfobins: double>
                child 0, capabilities_gtfobins: double
              avg_tokens_per_run: double
              traces_gt_16k: int64
              scenario_long_ctx: struct<capabilities_gtfobins: struct<16k: int64, 32k: int64, 64k: int64>>
                child 0, capabilities_gtfobins: struct<16k: int64, 32k: int64, 64k: int64>
                    child 0, 16k: int64
                    child 1, 32k: int64
                    child 2, 64k: int64
              min_turns: int64
              total_messages: int64
              avg_toolcalls_per_run: double
              max_tokens: int64
              max_turns: int64
              metadata_histograms: struct<generator_name: struct<capabilities_gtfobins: int64>, category: struct<capabilities_gtfobins: (... 153 chars omitted)
                child 0, generator_name: struct<capabilities_gtfobins: int64>
                    child 0, capabilities_gtfobins: int64
                child 1, category: struct<capabilities_gtfobins: int64>
                    child 0, capabilities_gtfobins: int64
                child 2, binary_name: struct<perl: int64, php: int64, ruby: int64>
                    child 0, perl: int64
                    child 1, php: int64
                    child 2, ruby: int64
                child 3, binary_path: struct</usr/bin/perl: int64, /usr/bin/php: int64, /usr/bin/ruby: int64>
                    child 0, /usr/bin/perl: int64
                    child 1, /usr/bin/php: int64
                    child 2, /usr/bin/ruby: int64
              toolcall_types: struct<exec_command: int64, test_credentials: int64>
                child 0, exec_command: int64
                child 1, test_credentials: int64
              avg_messages_per_run: double
              traces_gt_32k: int64
              min_tokens: int64
              total_toolcalls: int64
              to
              {'num_runs': Value('int64'), 'total_messages': Value('int64'), 'avg_messages_per_run': Value('float64'), 'total_toolcalls': Value('int64'), 'avg_toolcalls_per_run': Value('float64'), 'total_tokens': Value('int64'), 'avg_tokens_per_run': Value('float64'), 'scenarios': {'capabilities_gtfobins': Value('int64')}, 'toolcall_types': {'exec_command': Value('int64'), 'test_credentials': Value('int64')}, 'traces_gt_16k': Value('int64'), 'traces_gt_32k': Value('int64'), 'traces_gt_64k': Value('int64'), 'scenario_long_ctx': {'capabilities_gtfobins': {'16k': Value('int64'), '32k': Value('int64'), '64k': Value('int64')}}, 'scenario_avg_turns': {'capabilities_gtfobins': Value('float64')}, 'min_turns': Value('int64'), 'max_turns': Value('int64'), 'avg_turns': Value('float64'), 'min_tokens': Value('int64'), 'max_tokens': Value('int64'), 'metadata_histograms': {'generator_name': {'capabilities_gtfobins': Value('int64')}, 'category': {'capabilities_gtfobins': Value('int64')}, 'binary_name': {'perl': Value('int64'), 'php': Value('int64'), 'ruby': Value('int64')}, 'binary_path': {'/usr/bin/perl': Value('int64'), '/usr/bin/php': Value('int64'), '/usr/bin/ruby': Value('int64')}}}
              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.
PrivEsc-LLM Data

This dataset artifact contains the paper SFT dataset used for the final Qwen3-4B warm start, example transcripts, and the split/leakage audit.
Source repository: https://github.com/sailab-vienna/privesc-llm
Layout

paper_sft_dataset/: assembled training and validation JSONL traces for standard/unguided/deepseek/long_reasoning
leakage_audit/: generated split/leakage audit bundle with summary.json, split_holdouts.tsv, and benchmark_holdouts.tsv
examples/: 11 successful headline PrivEsc-LLM static-eval transcripts and one Docker failure transcript
Evaluation traces and summaries live in privesc-llm-evals. Credentials, usernames, host paths, and SSH keys appearing inside traces are synthetic benchmark artifacts from controlled local privilege-escalation scenarios.
Downloads last month: 63
Total file size:
75 MB