|
|
--- |
|
|
dataset_info: |
|
|
- config_name: AdvBench.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 3586281 |
|
|
num_examples: 1094 |
|
|
download_size: 1061590 |
|
|
dataset_size: 3586281 |
|
|
- config_name: CatQA.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1851591 |
|
|
num_examples: 550 |
|
|
download_size: 547045 |
|
|
dataset_size: 1851591 |
|
|
- config_name: CoCoNot.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 4216561 |
|
|
num_examples: 1380 |
|
|
download_size: 1302732 |
|
|
dataset_size: 4216561 |
|
|
- config_name: CoNA.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 880718 |
|
|
num_examples: 178 |
|
|
download_size: 220459 |
|
|
dataset_size: 880718 |
|
|
- config_name: CoSafe.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 5763461 |
|
|
num_examples: 1400 |
|
|
download_size: 1535748 |
|
|
dataset_size: 5763461 |
|
|
- config_name: ControversialInstructions.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 196094 |
|
|
num_examples: 40 |
|
|
download_size: 62890 |
|
|
dataset_size: 196094 |
|
|
- config_name: CyberattackAssistance.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 4754203 |
|
|
num_examples: 1000 |
|
|
download_size: 1160811 |
|
|
dataset_size: 4754203 |
|
|
- config_name: DAN.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1201428 |
|
|
num_examples: 390 |
|
|
download_size: 367379 |
|
|
dataset_size: 1201428 |
|
|
- config_name: DeMET.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 233894 |
|
|
num_examples: 29 |
|
|
download_size: 59615 |
|
|
dataset_size: 233894 |
|
|
- config_name: DiaSafety.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1735158 |
|
|
num_examples: 501 |
|
|
download_size: 581850 |
|
|
dataset_size: 1735158 |
|
|
- config_name: DoNotAnswer.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 3897191 |
|
|
num_examples: 939 |
|
|
download_size: 960375 |
|
|
dataset_size: 3897191 |
|
|
- config_name: ForbiddenQuestions.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1279124 |
|
|
num_examples: 390 |
|
|
download_size: 380109 |
|
|
dataset_size: 1279124 |
|
|
- config_name: GEST.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 12179045 |
|
|
num_examples: 3565 |
|
|
download_size: 3789943 |
|
|
dataset_size: 12179045 |
|
|
- config_name: GPTFuzzer.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 382675 |
|
|
num_examples: 100 |
|
|
download_size: 116691 |
|
|
dataset_size: 382675 |
|
|
- config_name: GandalfIgnoreInstructions.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 533568 |
|
|
num_examples: 112 |
|
|
download_size: 139573 |
|
|
dataset_size: 533568 |
|
|
- config_name: GandalfSummarization.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 97931 |
|
|
num_examples: 13 |
|
|
download_size: 39717 |
|
|
dataset_size: 97931 |
|
|
- config_name: HarmBench.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1253080 |
|
|
num_examples: 320 |
|
|
download_size: 365605 |
|
|
dataset_size: 1253080 |
|
|
- config_name: HarmfulQ.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 912684 |
|
|
num_examples: 200 |
|
|
download_size: 229225 |
|
|
dataset_size: 912684 |
|
|
- config_name: HarmfulQA.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 6547538 |
|
|
num_examples: 1960 |
|
|
download_size: 1905299 |
|
|
dataset_size: 6547538 |
|
|
- config_name: JADE.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 269611 |
|
|
num_examples: 80 |
|
|
download_size: 104471 |
|
|
dataset_size: 269611 |
|
|
- config_name: JBBBehaviours.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 285816 |
|
|
num_examples: 100 |
|
|
download_size: 108093 |
|
|
dataset_size: 285816 |
|
|
- config_name: LatentJailbreak.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 11355555 |
|
|
num_examples: 2424 |
|
|
download_size: 2320008 |
|
|
dataset_size: 11355555 |
|
|
- config_name: MaliciousInstruct.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 580525 |
|
|
num_examples: 100 |
|
|
download_size: 144707 |
|
|
dataset_size: 580525 |
|
|
- config_name: MaliciousInstructions.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 452430 |
|
|
num_examples: 100 |
|
|
download_size: 123879 |
|
|
dataset_size: 452430 |
|
|
- config_name: MedSafetyBench.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 3817597 |
|
|
num_examples: 900 |
|
|
download_size: 1082712 |
|
|
dataset_size: 3817597 |
|
|
- config_name: MoralExceptQA.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
struct: |
|
|
- name: human.response |
|
|
dtype: float64 |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 799604 |
|
|
num_examples: 148 |
|
|
download_size: 201400 |
|
|
dataset_size: 799604 |
|
|
- config_name: ORBench.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 4726479 |
|
|
num_examples: 1319 |
|
|
download_size: 1429290 |
|
|
dataset_size: 4726479 |
|
|
- config_name: PhysicalSafetyInstructions.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 761714 |
|
|
num_examples: 100 |
|
|
download_size: 176292 |
|
|
dataset_size: 761714 |
|
|
- config_name: QHarm.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 497451 |
|
|
num_examples: 100 |
|
|
download_size: 142196 |
|
|
dataset_size: 497451 |
|
|
- config_name: SGBench.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 4283265 |
|
|
num_examples: 1192 |
|
|
download_size: 1228895 |
|
|
dataset_size: 4283265 |
|
|
- config_name: SGXSTest.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 601806 |
|
|
num_examples: 100 |
|
|
download_size: 158946 |
|
|
dataset_size: 601806 |
|
|
- config_name: SafeText.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
struct: |
|
|
- name: safe |
|
|
sequence: string |
|
|
- name: unsafe |
|
|
sequence: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1960023 |
|
|
num_examples: 367 |
|
|
download_size: 576260 |
|
|
dataset_size: 1960023 |
|
|
- config_name: StrongREJECT.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1397926 |
|
|
num_examples: 313 |
|
|
download_size: 396184 |
|
|
dataset_size: 1397926 |
|
|
- config_name: ToxiGen.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 3033249 |
|
|
num_examples: 940 |
|
|
download_size: 1022508 |
|
|
dataset_size: 3033249 |
|
|
- config_name: WMDP.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
sequence: string |
|
|
- name: answer |
|
|
dtype: int64 |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 19110435 |
|
|
num_examples: 3668 |
|
|
download_size: 5360635 |
|
|
dataset_size: 19110435 |
|
|
- config_name: XSTest.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 1526876 |
|
|
num_examples: 450 |
|
|
download_size: 467338 |
|
|
dataset_size: 1526876 |
|
|
- config_name: XSafety.json |
|
|
features: |
|
|
- name: prompt |
|
|
dtype: string |
|
|
- name: choices |
|
|
dtype: string |
|
|
- name: answer |
|
|
dtype: string |
|
|
- name: task |
|
|
dtype: string |
|
|
- name: subtask |
|
|
dtype: string |
|
|
- name: category |
|
|
dtype: string |
|
|
- name: domain |
|
|
dtype: string |
|
|
- name: language |
|
|
dtype: string |
|
|
- name: source |
|
|
dtype: string |
|
|
- name: risk_response |
|
|
dtype: string |
|
|
- name: risk_property |
|
|
dtype: string |
|
|
- name: domain_response |
|
|
dtype: string |
|
|
- name: domain_property |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 9463190 |
|
|
num_examples: 2800 |
|
|
download_size: 2940828 |
|
|
dataset_size: 9463190 |
|
|
configs: |
|
|
- config_name: AdvBench.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: AdvBench.json/train-* |
|
|
- config_name: CatQA.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: CatQA.json/train-* |
|
|
- config_name: CoCoNot.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: CoCoNot.json/train-* |
|
|
- config_name: CoNA.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: CoNA.json/train-* |
|
|
- config_name: CoSafe.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: CoSafe.json/train-* |
|
|
- config_name: ControversialInstructions.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: ControversialInstructions.json/train-* |
|
|
- config_name: CyberattackAssistance.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: CyberattackAssistance.json/train-* |
|
|
- config_name: DAN.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: DAN.json/train-* |
|
|
- config_name: DeMET.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: DeMET.json/train-* |
|
|
- config_name: DiaSafety.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: DiaSafety.json/train-* |
|
|
- config_name: DoNotAnswer.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: DoNotAnswer.json/train-* |
|
|
- config_name: ForbiddenQuestions.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: ForbiddenQuestions.json/train-* |
|
|
- config_name: GEST.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: GEST.json/train-* |
|
|
- config_name: GPTFuzzer.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: GPTFuzzer.json/train-* |
|
|
- config_name: GandalfIgnoreInstructions.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: GandalfIgnoreInstructions.json/train-* |
|
|
- config_name: GandalfSummarization.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: GandalfSummarization.json/train-* |
|
|
- config_name: HarmBench.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: HarmBench.json/train-* |
|
|
- config_name: HarmfulQ.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: HarmfulQ.json/train-* |
|
|
- config_name: HarmfulQA.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: HarmfulQA.json/train-* |
|
|
- config_name: JADE.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: JADE.json/train-* |
|
|
- config_name: JBBBehaviours.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: JBBBehaviours.json/train-* |
|
|
- config_name: LatentJailbreak.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: LatentJailbreak.json/train-* |
|
|
- config_name: MaliciousInstruct.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: MaliciousInstruct.json/train-* |
|
|
- config_name: MaliciousInstructions.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: MaliciousInstructions.json/train-* |
|
|
- config_name: MedSafetyBench.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: MedSafetyBench.json/train-* |
|
|
- config_name: MoralExceptQA.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: MoralExceptQA.json/train-* |
|
|
- config_name: ORBench.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: ORBench.json/train-* |
|
|
- config_name: PhysicalSafetyInstructions.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: PhysicalSafetyInstructions.json/train-* |
|
|
- config_name: QHarm.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: QHarm.json/train-* |
|
|
- config_name: SGBench.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: SGBench.json/train-* |
|
|
- config_name: SGXSTest.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: SGXSTest.json/train-* |
|
|
- config_name: SafeText.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: SafeText.json/train-* |
|
|
- config_name: StrongREJECT.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: StrongREJECT.json/train-* |
|
|
- config_name: ToxiGen.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: ToxiGen.json/train-* |
|
|
- config_name: WMDP.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: WMDP.json/train-* |
|
|
- config_name: XSTest.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: XSTest.json/train-* |
|
|
- config_name: XSafety.json |
|
|
data_files: |
|
|
- split: train |
|
|
path: XSafety.json/train-* |
|
|
--- |
|
|
|
