| | --- |
| | dataset_info: |
| | features: |
| | - name: action_id |
| | dtype: int64 |
| | - name: context |
| | dtype: string |
| | - name: instruction |
| | dtype: string |
| | - name: initialization |
| | struct: |
| | - name: context |
| | dtype: string |
| | - name: id |
| | dtype: int64 |
| | - name: initialization |
| | struct: |
| | - name: difficulty |
| | dtype: int64 |
| | - name: model |
| | dtype: string |
| | - name: original_trajectory_messages |
| | list: |
| | - name: content |
| | dtype: string |
| | - name: role |
| | dtype: string |
| | - name: original_trajectory_tag |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: src_idx |
| | dtype: int64 |
| | - name: intended_classification |
| | dtype: string |
| | - name: mode |
| | dtype: string |
| | - name: auxiliary_outputs |
| | struct: |
| | - name: prompt_for_monitors |
| | dtype: string |
| | - name: prompt_with_no_trigger |
| | dtype: string |
| | - name: prompt |
| | dtype: string |
| | - name: prompt_that_generated_output |
| | dtype: string |
| | - name: classification |
| | dtype: string |
| | - name: output |
| | dtype: string |
| | - name: prompt_for_grader |
| | dtype: string |
| | - name: graders |
| | struct: |
| | - name: gpt-4o |
| | struct: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: messages |
| | list: |
| | - name: content |
| | dtype: string |
| | - name: role |
| | dtype: string |
| | - name: monitors |
| | struct: |
| | - name: qwen-2.5-32b-coder |
| | list: |
| | list: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: monitor_explanation_correct |
| | struct: |
| | - name: assessment |
| | dtype: bool |
| | - name: grader |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: prompt_prefix |
| | dtype: string |
| | - name: off_policy |
| | dtype: bool |
| | - name: includes_trigger |
| | dtype: bool |
| | - name: generating_model |
| | dtype: string |
| | - name: action_kept |
| | dtype: bool |
| | - name: complexity |
| | dtype: string |
| | - name: trajectory_kept |
| | dtype: bool |
| | - name: trajectory_index |
| | dtype: int64 |
| | - name: validation |
| | struct: |
| | - name: error_type |
| | dtype: 'null' |
| | - name: gpt_4o |
| | struct: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: raw_text |
| | dtype: string |
| | - name: gpt_4o_mini_1 |
| | struct: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: raw_text |
| | dtype: string |
| | - name: gpt_4o_mini_2 |
| | struct: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: raw_text |
| | dtype: string |
| | - name: had_disagreement |
| | dtype: bool |
| | - name: intended_classification |
| | dtype: string |
| | - name: logical_disagreement_details |
| | struct: |
| | - name: intended |
| | dtype: string |
| | - name: llm_classifications |
| | sequence: string |
| | - name: occurred |
| | dtype: bool |
| | - name: prompt_for_validation |
| | dtype: string |
| | - name: action_one |
| | struct: |
| | - name: action_id |
| | dtype: int64 |
| | - name: action_kept |
| | dtype: bool |
| | - name: auxiliary_outputs |
| | struct: |
| | - name: prompt_for_monitors |
| | dtype: string |
| | - name: classification |
| | dtype: string |
| | - name: complexity |
| | dtype: string |
| | - name: context |
| | dtype: string |
| | - name: generating_model |
| | dtype: string |
| | - name: graders |
| | struct: |
| | - name: gpt-4o |
| | struct: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: includes_trigger |
| | dtype: bool |
| | - name: initialization |
| | struct: |
| | - name: context |
| | dtype: string |
| | - name: id |
| | dtype: int64 |
| | - name: initialization |
| | struct: |
| | - name: difficulty |
| | dtype: int64 |
| | - name: model |
| | dtype: string |
| | - name: original_trajectory_messages |
| | list: |
| | - name: content |
| | dtype: string |
| | - name: role |
| | dtype: string |
| | - name: original_trajectory_tag |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: src_idx |
| | dtype: int64 |
| | - name: instruction |
| | dtype: string |
| | - name: intended_classification |
| | dtype: string |
| | - name: messages |
| | list: |
| | - name: content |
| | dtype: string |
| | - name: role |
| | dtype: string |
| | - name: mode |
| | dtype: string |
| | - name: monitors |
| | struct: |
| | - name: qwen-2.5-32b-coder |
| | list: |
| | list: |
| | - name: classification |
| | dtype: string |
| | - name: explanation |
| | dtype: string |
| | - name: monitor_explanation_correct |
| | struct: |
| | - name: assessment |
| | dtype: bool |
| | - name: grader |
| | dtype: string |
| | - name: thoughts |
| | dtype: string |
| | - name: prompt_prefix |
| | dtype: string |
| | - name: off_policy |
| | dtype: bool |
| | - name: output |
| | dtype: string |
| | - name: prompt |
| | dtype: string |
| | - name: prompt_for_grader |
| | dtype: string |
| | - name: prompt_that_generated_output |
| | dtype: string |
| | - name: prompt_with_no_trigger |
| | dtype: string |
| | - name: trajectory_index |
| | dtype: int64 |
| | - name: trajectory_kept |
| | dtype: bool |
| | splits: |
| | - name: train |
| | num_bytes: 7730590 |
| | num_examples: 100 |
| | download_size: 2334347 |
| | dataset_size: 7730590 |
| | configs: |
| | - config_name: default |
| | data_files: |
| | - split: train |
| | path: data/train-* |
| | --- |
| | |