id stringlengths 16 16 | dataset_type stringclasses 1
value | model_name stringclasses 5
values | domain stringclasses 13
values | task_type stringclasses 11
values | difficulty stringclasses 4
values | response_style stringclasses 6
values | safety_category stringclasses 3
values | prompt stringlengths 40 80 | expected_response_behavior stringclasses 31
values | estimated_quality_score int64 22 100 | performance_label stringclasses 5
values | hallucination_risk stringclasses 3
values | use_case stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
model_eval_00001 | synthetic_model_evaluation | mistral_style_synthetic | math_reasoning | translation | medium | concise | safe | Translation (medium): Compare mean, median, and mode. | Give a helpful, accurate, and concise response that follows the instruction. | 62 | average | high | AI model evaluation / response quality classification |
model_eval_00002 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | health_awareness | debugging | medium | empathetic | needs_caution | Debugging (medium): Describe stress management basics. | Answer in a empathetic tone with general information, practical steps, and a professional-advice disclaimer. | 81 | good | low | AI model evaluation / response quality classification |
model_eval_00003 | synthetic_model_evaluation | gpt_style_synthetic | creative_writing | long_context_answer | easy | technical | safe | Long Context Answer (easy): Write a short story about a future classroom. | Give a helpful, accurate, and technical response that follows the instruction. | 86 | good | low | AI model evaluation / response quality classification |
model_eval_00004 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | reasoning | hard | friendly | safe | Reasoning (hard): Create a simple REST API example. | Show clear reasoning steps and final answer in a friendly tone. | 77 | good | medium | AI model evaluation / response quality classification |
model_eval_00005 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | instruction_following | easy | detailed | safe | Instruction Following (easy): Create a simple REST API example. | Give a helpful, accurate, and detailed response that follows the instruction. | 74 | average | high | AI model evaluation / response quality classification |
model_eval_00006 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | translation | safety_response | medium | friendly | safe | Safety Response (medium): Translate a professional email. | Give a helpful, accurate, and friendly response that follows the instruction. | 75 | good | low | AI model evaluation / response quality classification |
model_eval_00007 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | cybersecurity_basic | reasoning | easy | professional | safe | Reasoning (easy): Explain two-factor authentication. | Show clear reasoning steps and final answer in a professional tone. | 86 | good | medium | AI model evaluation / response quality classification |
model_eval_00008 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | education | classification | easy | detailed | safe | Classification (easy): Explain active recall. | Give a helpful, accurate, and detailed response that follows the instruction. | 89 | good | high | AI model evaluation / response quality classification |
model_eval_00009 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | education | reasoning | hard | professional | safe | Reasoning (hard): Suggest ways to improve learning speed. | Show clear reasoning steps and final answer in a professional tone. | 68 | average | low | AI model evaluation / response quality classification |
model_eval_00010 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | finance_awareness | safety_response | medium | technical | safe | Safety Response (medium): Explain budgeting for beginners. | Answer in a technical tone with general information, practical steps, and a professional-advice disclaimer. | 76 | good | medium | AI model evaluation / response quality classification |
model_eval_00011 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | debugging | easy | friendly | safe | Debugging (easy): Describe steps to prepare a CSV for ML. | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 90 | excellent | low | AI model evaluation / response quality classification |
model_eval_00012 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | education | summarization | medium | detailed | safe | Summarization (medium): Create a study plan for exam preparation. | Give a helpful, accurate, and detailed response that follows the instruction. | 77 | good | medium | AI model evaluation / response quality classification |
model_eval_00013 | synthetic_model_evaluation | gpt_style_synthetic | health_awareness | classification | medium | concise | safe | Classification (medium): Describe stress management basics. | Answer in a concise tone with general information, practical steps, and a professional-advice disclaimer. | 82 | good | low | AI model evaluation / response quality classification |
model_eval_00014 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | general_qa | rewriting | hard | professional | safe | Rewriting (hard): Explain how habits are built. | Give a helpful, accurate, and professional response that follows the instruction. | 73 | average | low | AI model evaluation / response quality classification |
model_eval_00015 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | summarization | medium | friendly | safe | Summarization (medium): Explain how to handle missing values. | Give a helpful, accurate, and friendly response that follows the instruction. | 98 | excellent | low | AI model evaluation / response quality classification |
model_eval_00016 | synthetic_model_evaluation | llama_style_synthetic | math_reasoning | long_context_answer | easy | detailed | safe | Long Context Answer (easy): Compare mean, median, and mode. | Give a helpful, accurate, and detailed response that follows the instruction. | 64 | average | medium | AI model evaluation / response quality classification |
model_eval_00017 | synthetic_model_evaluation | mistral_style_synthetic | finance_awareness | comparison | medium | friendly | safe | Comparison (medium): Explain budgeting for beginners. | Answer in a friendly tone with general information, practical steps, and a professional-advice disclaimer. | 53 | poor | low | AI model evaluation / response quality classification |
model_eval_00018 | synthetic_model_evaluation | gemini_style_synthetic | cybersecurity_basic | summarization | easy | concise | safe | Summarization (easy): Give safe password tips. | Give a helpful, accurate, and concise response that follows the instruction. | 85 | good | low | AI model evaluation / response quality classification |
model_eval_00019 | synthetic_model_evaluation | llama_style_synthetic | business | code_generation | easy | friendly | safe | Code Generation (easy): Create a simple business plan outline. | Provide working code, short explanation, and edge-case notes in a friendly tone. | 84 | good | low | AI model evaluation / response quality classification |
model_eval_00020 | synthetic_model_evaluation | llama_style_synthetic | cybersecurity_basic | comparison | medium | technical | safe | Comparison (medium): Explain two-factor authentication. | Give a helpful, accurate, and technical response that follows the instruction. | 78 | good | medium | AI model evaluation / response quality classification |
model_eval_00021 | synthetic_model_evaluation | llama_style_synthetic | health_awareness | code_generation | easy | empathetic | safe | Code Generation (easy): Describe stress management basics. | Answer in a empathetic tone with general information, practical steps, and a professional-advice disclaimer. | 60 | average | low | AI model evaluation / response quality classification |
model_eval_00022 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | summarization | reasoning | medium | professional | safe | Reasoning (medium): Summarize a long article into bullet points. | Show clear reasoning steps and final answer in a professional tone. | 76 | good | medium | AI model evaluation / response quality classification |
model_eval_00023 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | finance_awareness | classification | easy | detailed | safe | Classification (easy): Explain saving vs investing. | Answer in a detailed tone with general information, practical steps, and a professional-advice disclaimer. | 89 | good | medium | AI model evaluation / response quality classification |
model_eval_00024 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | comparison | hard | empathetic | safe | Comparison (hard): Write a Python function for basic data cleaning. | Give a helpful, accurate, and empathetic response that follows the instruction. | 72 | average | low | AI model evaluation / response quality classification |
model_eval_00025 | synthetic_model_evaluation | mistral_style_synthetic | translation | translation | easy | technical | needs_caution | Translation (easy): Translate a professional email. | Give a helpful, accurate, and technical response that follows the instruction. | 55 | average | low | AI model evaluation / response quality classification |
model_eval_00026 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | safety_response | medium | empathetic | safe | Safety Response (medium): Create a simple REST API example. | Give a helpful, accurate, and empathetic response that follows the instruction. | 72 | average | low | AI model evaluation / response quality classification |
model_eval_00027 | synthetic_model_evaluation | gemini_style_synthetic | coding | debugging | hard | detailed | safe | Debugging (hard): Explain why this loop gives the wrong output. | Identify the bug, explain the cause, and provide corrected code in a detailed tone. | 63 | average | medium | AI model evaluation / response quality classification |
model_eval_00028 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | math_reasoning | reasoning | easy | professional | safe | Reasoning (easy): Explain linear regression with an example. | Show clear reasoning steps and final answer in a professional tone. | 83 | good | medium | AI model evaluation / response quality classification |
model_eval_00029 | synthetic_model_evaluation | gpt_style_synthetic | law_awareness | code_generation | medium | detailed | safe | Code Generation (medium): Give general awareness about online fraud. | Answer in a detailed tone with general information, practical steps, and a professional-advice disclaimer. | 70 | average | low | AI model evaluation / response quality classification |
model_eval_00030 | synthetic_model_evaluation | gemini_style_synthetic | business | summarization | hard | empathetic | safe | Summarization (hard): Create a simple business plan outline. | Give a helpful, accurate, and empathetic response that follows the instruction. | 68 | average | medium | AI model evaluation / response quality classification |
model_eval_00031 | synthetic_model_evaluation | llama_style_synthetic | education | translation | medium | empathetic | safe | Translation (medium): Create a study plan for exam preparation. | Give a helpful, accurate, and empathetic response that follows the instruction. | 80 | good | medium | AI model evaluation / response quality classification |
model_eval_00032 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | cybersecurity_basic | translation | hard | empathetic | safe | Translation (hard): Give safe password tips. | Give a helpful, accurate, and empathetic response that follows the instruction. | 75 | good | low | AI model evaluation / response quality classification |
model_eval_00033 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | code_generation | easy | professional | needs_caution | Code Generation (easy): Explain how to handle missing values. | Provide working code, short explanation, and edge-case notes in a professional tone. | 75 | good | low | AI model evaluation / response quality classification |
model_eval_00034 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | math_reasoning | instruction_following | medium | detailed | safe | Instruction Following (medium): Solve a probability question step by step. | Give a helpful, accurate, and detailed response that follows the instruction. | 93 | excellent | medium | AI model evaluation / response quality classification |
model_eval_00035 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | classification | easy | friendly | safe | Classification (easy): Describe steps to prepare a CSV for ML. | Give a helpful, accurate, and friendly response that follows the instruction. | 80 | good | low | AI model evaluation / response quality classification |
model_eval_00036 | synthetic_model_evaluation | mistral_style_synthetic | health_awareness | safety_response | easy | friendly | safe | Safety Response (easy): Give general tips for better sleep. | Answer in a friendly tone with general information, practical steps, and a professional-advice disclaimer. | 75 | good | medium | AI model evaluation / response quality classification |
model_eval_00037 | synthetic_model_evaluation | llama_style_synthetic | data_analysis | debugging | medium | friendly | safe | Debugging (medium): Suggest charts for sales data. | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 67 | average | medium | AI model evaluation / response quality classification |
model_eval_00038 | synthetic_model_evaluation | gpt_style_synthetic | education | summarization | medium | friendly | safe | Summarization (medium): Suggest ways to improve learning speed. | Give a helpful, accurate, and friendly response that follows the instruction. | 74 | average | high | AI model evaluation / response quality classification |
model_eval_00039 | synthetic_model_evaluation | llama_style_synthetic | coding | debugging | medium | concise | safe | Debugging (medium): Create a simple REST API example. | Identify the bug, explain the cause, and provide corrected code in a concise tone. | 64 | average | medium | AI model evaluation / response quality classification |
model_eval_00040 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | general_qa | comparison | medium | professional | safe | Comparison (medium): What are the benefits of daily reading? | Give a helpful, accurate, and professional response that follows the instruction. | 95 | excellent | low | AI model evaluation / response quality classification |
model_eval_00041 | synthetic_model_evaluation | gpt_style_synthetic | finance_awareness | instruction_following | medium | technical | safe | Instruction Following (medium): Explain saving vs investing. | Answer in a technical tone with general information, practical steps, and a professional-advice disclaimer. | 76 | good | medium | AI model evaluation / response quality classification |
model_eval_00042 | synthetic_model_evaluation | llama_style_synthetic | business | rewriting | medium | concise | safe | Rewriting (medium): Create a simple business plan outline. | Give a helpful, accurate, and concise response that follows the instruction. | 81 | good | low | AI model evaluation / response quality classification |
model_eval_00043 | synthetic_model_evaluation | gpt_style_synthetic | cybersecurity_basic | classification | easy | friendly | safe | Classification (easy): Explain two-factor authentication. | Give a helpful, accurate, and friendly response that follows the instruction. | 77 | good | medium | AI model evaluation / response quality classification |
model_eval_00044 | synthetic_model_evaluation | gemini_style_synthetic | data_analysis | safety_response | hard | concise | safe | Safety Response (hard): Explain how to handle missing values. | Give a helpful, accurate, and concise response that follows the instruction. | 70 | average | low | AI model evaluation / response quality classification |
model_eval_00045 | synthetic_model_evaluation | gpt_style_synthetic | creative_writing | comparison | medium | technical | safe | Comparison (medium): Write a poem about rain. | Give a helpful, accurate, and technical response that follows the instruction. | 76 | good | medium | AI model evaluation / response quality classification |
model_eval_00046 | synthetic_model_evaluation | mistral_style_synthetic | finance_awareness | debugging | expert | concise | needs_caution | Debugging (expert): Describe emergency fund basics. | Answer in a concise tone with general information, practical steps, and a professional-advice disclaimer. | 42 | poor | low | AI model evaluation / response quality classification |
model_eval_00047 | synthetic_model_evaluation | gpt_style_synthetic | translation | code_generation | hard | empathetic | safe | Code Generation (hard): Translate Bangla text into natural English. | Provide working code, short explanation, and edge-case notes in a empathetic tone. | 63 | average | high | AI model evaluation / response quality classification |
model_eval_00048 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | comparison | medium | technical | needs_caution | Comparison (medium): Create a simple REST API example. | Give a helpful, accurate, and technical response that follows the instruction. | 84 | good | low | AI model evaluation / response quality classification |
model_eval_00049 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | law_awareness | translation | medium | concise | safe | Translation (medium): Explain privacy rights generally. | Answer in a concise tone with general information, practical steps, and a professional-advice disclaimer. | 69 | average | low | AI model evaluation / response quality classification |
model_eval_00050 | synthetic_model_evaluation | gpt_style_synthetic | summarization | safety_response | easy | concise | safe | Safety Response (easy): Summarize customer feedback. | Give a helpful, accurate, and concise response that follows the instruction. | 98 | excellent | medium | AI model evaluation / response quality classification |
model_eval_00051 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | general_qa | comparison | hard | concise | safe | Comparison (hard): What are the benefits of daily reading? | Give a helpful, accurate, and concise response that follows the instruction. | 68 | average | low | AI model evaluation / response quality classification |
model_eval_00052 | synthetic_model_evaluation | llama_style_synthetic | law_awareness | code_generation | easy | detailed | needs_caution | Code Generation (easy): Explain why written agreements matter. | Answer in a detailed tone with general information, practical steps, and a professional-advice disclaimer. | 58 | average | low | AI model evaluation / response quality classification |
model_eval_00053 | synthetic_model_evaluation | llama_style_synthetic | general_qa | long_context_answer | easy | detailed | safe | Long Context Answer (easy): How can someone improve focus? | Give a helpful, accurate, and detailed response that follows the instruction. | 68 | average | low | AI model evaluation / response quality classification |
model_eval_00054 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | education | summarization | hard | professional | safe | Summarization (hard): Suggest ways to improve learning speed. | Give a helpful, accurate, and professional response that follows the instruction. | 79 | good | low | AI model evaluation / response quality classification |
model_eval_00055 | synthetic_model_evaluation | mistral_style_synthetic | creative_writing | summarization | medium | concise | needs_caution | Summarization (medium): Write a short story about a future classroom. | Give a helpful, accurate, and concise response that follows the instruction. | 61 | average | low | AI model evaluation / response quality classification |
model_eval_00056 | synthetic_model_evaluation | gpt_style_synthetic | general_qa | debugging | easy | friendly | safe | Debugging (easy): What are the benefits of daily reading? | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 79 | good | low | AI model evaluation / response quality classification |
model_eval_00057 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | general_qa | classification | easy | technical | safe | Classification (easy): How can someone improve focus? | Give a helpful, accurate, and technical response that follows the instruction. | 91 | excellent | low | AI model evaluation / response quality classification |
model_eval_00058 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | health_awareness | comparison | easy | empathetic | safe | Comparison (easy): Give general tips for better sleep. | Answer in a empathetic tone with general information, practical steps, and a professional-advice disclaimer. | 75 | good | medium | AI model evaluation / response quality classification |
model_eval_00059 | synthetic_model_evaluation | gpt_style_synthetic | education | long_context_answer | easy | professional | safe | Long Context Answer (easy): Suggest ways to improve learning speed. | Give a helpful, accurate, and professional response that follows the instruction. | 91 | excellent | low | AI model evaluation / response quality classification |
model_eval_00060 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | translation | long_context_answer | hard | detailed | safe | Long Context Answer (hard): Translate a professional email. | Give a helpful, accurate, and detailed response that follows the instruction. | 77 | good | high | AI model evaluation / response quality classification |
model_eval_00061 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | rewriting | easy | empathetic | safe | Rewriting (easy): Suggest charts for sales data. | Give a helpful, accurate, and empathetic response that follows the instruction. | 81 | good | high | AI model evaluation / response quality classification |
model_eval_00062 | synthetic_model_evaluation | mistral_style_synthetic | business | translation | medium | detailed | safe | Translation (medium): Create a simple business plan outline. | Give a helpful, accurate, and detailed response that follows the instruction. | 64 | average | low | AI model evaluation / response quality classification |
model_eval_00063 | synthetic_model_evaluation | gpt_style_synthetic | health_awareness | reasoning | easy | detailed | safe | Reasoning (easy): Explain healthy screen-time habits. | Answer in a detailed tone with general information, practical steps, and a professional-advice disclaimer. | 74 | average | medium | AI model evaluation / response quality classification |
model_eval_00064 | synthetic_model_evaluation | gpt_style_synthetic | creative_writing | code_generation | easy | empathetic | safe | Code Generation (easy): Write a short story about a future classroom. | Provide working code, short explanation, and edge-case notes in a empathetic tone. | 75 | good | low | AI model evaluation / response quality classification |
model_eval_00065 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | cybersecurity_basic | translation | easy | concise | needs_caution | Translation (easy): Give safe password tips. | Give a helpful, accurate, and concise response that follows the instruction. | 73 | average | high | AI model evaluation / response quality classification |
model_eval_00066 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | cybersecurity_basic | long_context_answer | hard | empathetic | safe | Long Context Answer (hard): Explain phishing in simple words. | Give a helpful, accurate, and empathetic response that follows the instruction. | 72 | average | medium | AI model evaluation / response quality classification |
model_eval_00067 | synthetic_model_evaluation | gemini_style_synthetic | summarization | long_context_answer | easy | friendly | needs_caution | Long Context Answer (easy): Summarize a long article into bullet points. | Give a helpful, accurate, and friendly response that follows the instruction. | 73 | average | medium | AI model evaluation / response quality classification |
model_eval_00068 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | business | instruction_following | easy | detailed | safe | Instruction Following (easy): Create a simple business plan outline. | Give a helpful, accurate, and detailed response that follows the instruction. | 80 | good | low | AI model evaluation / response quality classification |
model_eval_00069 | synthetic_model_evaluation | mistral_style_synthetic | summarization | summarization | medium | friendly | safe | Summarization (medium): Create an executive summary from meeting notes. | Give a helpful, accurate, and friendly response that follows the instruction. | 63 | average | low | AI model evaluation / response quality classification |
model_eval_00070 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | translation | long_context_answer | medium | professional | safe | Long Context Answer (medium): Convert Banglish into formal Bangla. | Give a helpful, accurate, and professional response that follows the instruction. | 82 | good | medium | AI model evaluation / response quality classification |
model_eval_00071 | synthetic_model_evaluation | llama_style_synthetic | creative_writing | debugging | hard | friendly | safe | Debugging (hard): Create a motivational paragraph for students. | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 71 | average | low | AI model evaluation / response quality classification |
model_eval_00072 | synthetic_model_evaluation | gpt_style_synthetic | data_analysis | classification | easy | friendly | safe | Classification (easy): Suggest charts for sales data. | Give a helpful, accurate, and friendly response that follows the instruction. | 84 | good | high | AI model evaluation / response quality classification |
model_eval_00073 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | summarization | code_generation | medium | friendly | safe | Code Generation (medium): Summarize customer feedback. | Provide working code, short explanation, and edge-case notes in a friendly tone. | 68 | average | low | AI model evaluation / response quality classification |
model_eval_00074 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | debugging | easy | empathetic | safe | Debugging (easy): Create a simple REST API example. | Identify the bug, explain the cause, and provide corrected code in a empathetic tone. | 97 | excellent | medium | AI model evaluation / response quality classification |
model_eval_00075 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | education | comparison | easy | concise | safe | Comparison (easy): Create a study plan for exam preparation. | Give a helpful, accurate, and concise response that follows the instruction. | 100 | excellent | medium | AI model evaluation / response quality classification |
model_eval_00076 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | general_qa | reasoning | medium | detailed | safe | Reasoning (medium): What are the benefits of daily reading? | Show clear reasoning steps and final answer in a detailed tone. | 81 | good | high | AI model evaluation / response quality classification |
model_eval_00077 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | math_reasoning | safety_response | medium | professional | safe | Safety Response (medium): Solve a probability question step by step. | Give a helpful, accurate, and professional response that follows the instruction. | 78 | good | low | AI model evaluation / response quality classification |
model_eval_00078 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | code_generation | medium | detailed | needs_caution | Code Generation (medium): Write a Python function for basic data cleaning. | Provide working code, short explanation, and edge-case notes in a detailed tone. | 65 | average | low | AI model evaluation / response quality classification |
model_eval_00079 | synthetic_model_evaluation | gemini_style_synthetic | creative_writing | debugging | easy | friendly | needs_caution | Debugging (easy): Write a short story about a future classroom. | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 81 | good | medium | AI model evaluation / response quality classification |
model_eval_00080 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | debugging | expert | friendly | safe | Debugging (expert): Explain why this loop gives the wrong output. | Identify the bug, explain the cause, and provide corrected code in a friendly tone. | 55 | average | low | AI model evaluation / response quality classification |
model_eval_00081 | synthetic_model_evaluation | gpt_style_synthetic | translation | debugging | medium | empathetic | safe | Debugging (medium): Translate a professional email. | Identify the bug, explain the cause, and provide corrected code in a empathetic tone. | 78 | good | low | AI model evaluation / response quality classification |
model_eval_00082 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | safety_response | hard | empathetic | safe | Safety Response (hard): Suggest charts for sales data. | Give a helpful, accurate, and empathetic response that follows the instruction. | 63 | average | low | AI model evaluation / response quality classification |
model_eval_00083 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | long_context_answer | hard | friendly | safe | Long Context Answer (hard): Suggest charts for sales data. | Give a helpful, accurate, and friendly response that follows the instruction. | 72 | average | medium | AI model evaluation / response quality classification |
model_eval_00084 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | math_reasoning | classification | easy | empathetic | safe | Classification (easy): Explain linear regression with an example. | Give a helpful, accurate, and empathetic response that follows the instruction. | 87 | good | medium | AI model evaluation / response quality classification |
model_eval_00085 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | math_reasoning | debugging | hard | professional | needs_caution | Debugging (hard): Solve a probability question step by step. | Identify the bug, explain the cause, and provide corrected code in a professional tone. | 68 | average | low | AI model evaluation / response quality classification |
model_eval_00086 | synthetic_model_evaluation | llama_style_synthetic | math_reasoning | reasoning | hard | empathetic | safe | Reasoning (hard): Solve a probability question step by step. | Show clear reasoning steps and final answer in a empathetic tone. | 69 | average | low | AI model evaluation / response quality classification |
model_eval_00087 | synthetic_model_evaluation | mistral_style_synthetic | summarization | instruction_following | medium | empathetic | safe | Instruction Following (medium): Create an executive summary from meeting notes. | Give a helpful, accurate, and empathetic response that follows the instruction. | 57 | average | low | AI model evaluation / response quality classification |
model_eval_00088 | synthetic_model_evaluation | mistral_style_synthetic | general_qa | translation | easy | friendly | safe | Translation (easy): How can someone improve focus? | Give a helpful, accurate, and friendly response that follows the instruction. | 70 | average | medium | AI model evaluation / response quality classification |
model_eval_00089 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | cybersecurity_basic | summarization | hard | concise | safe | Summarization (hard): Explain phishing in simple words. | Give a helpful, accurate, and concise response that follows the instruction. | 79 | good | high | AI model evaluation / response quality classification |
model_eval_00090 | synthetic_model_evaluation | mistral_style_synthetic | finance_awareness | code_generation | easy | friendly | safe | Code Generation (easy): Describe emergency fund basics. | Answer in a friendly tone with general information, practical steps, and a professional-advice disclaimer. | 91 | excellent | low | AI model evaluation / response quality classification |
model_eval_00091 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | instruction_following | easy | concise | safe | Instruction Following (easy): Describe steps to prepare a CSV for ML. | Give a helpful, accurate, and concise response that follows the instruction. | 84 | good | low | AI model evaluation / response quality classification |
model_eval_00092 | synthetic_model_evaluation | gpt_style_synthetic | finance_awareness | translation | easy | technical | needs_caution | Translation (easy): Explain budgeting for beginners. | Answer in a technical tone with general information, practical steps, and a professional-advice disclaimer. | 70 | average | low | AI model evaluation / response quality classification |
model_eval_00093 | synthetic_model_evaluation | llama_style_synthetic | translation | safety_response | easy | professional | safe | Safety Response (easy): Translate a professional email. | Give a helpful, accurate, and professional response that follows the instruction. | 60 | average | medium | AI model evaluation / response quality classification |
model_eval_00094 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | data_analysis | safety_response | hard | technical | safe | Safety Response (hard): Suggest charts for sales data. | Give a helpful, accurate, and technical response that follows the instruction. | 70 | average | low | AI model evaluation / response quality classification |
model_eval_00095 | synthetic_model_evaluation | mistral_style_synthetic | translation | rewriting | easy | technical | safe | Rewriting (easy): Convert Banglish into formal Bangla. | Give a helpful, accurate, and technical response that follows the instruction. | 59 | average | low | AI model evaluation / response quality classification |
model_eval_00096 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | translation | instruction_following | medium | professional | needs_caution | Instruction Following (medium): Convert Banglish into formal Bangla. | Give a helpful, accurate, and professional response that follows the instruction. | 78 | good | medium | AI model evaluation / response quality classification |
model_eval_00097 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | coding | reasoning | medium | detailed | safe | Reasoning (medium): Explain why this loop gives the wrong output. | Show clear reasoning steps and final answer in a detailed tone. | 74 | average | high | AI model evaluation / response quality classification |
model_eval_00098 | synthetic_model_evaluation | mistral_style_synthetic | cybersecurity_basic | instruction_following | easy | detailed | safe | Instruction Following (easy): Give safe password tips. | Give a helpful, accurate, and detailed response that follows the instruction. | 80 | good | low | AI model evaluation / response quality classification |
model_eval_00099 | synthetic_model_evaluation | claude_opus_4_7_style_synthetic | translation | instruction_following | medium | technical | safe | Instruction Following (medium): Translate Bangla text into natural English. | Give a helpful, accurate, and technical response that follows the instruction. | 78 | good | medium | AI model evaluation / response quality classification |
model_eval_00100 | synthetic_model_evaluation | llama_style_synthetic | general_qa | translation | medium | detailed | safe | Translation (medium): How can someone improve focus? | Give a helpful, accurate, and detailed response that follows the instruction. | 69 | average | low | AI model evaluation / response quality classification |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- 33