MathMate / eval /guardrail_cases.json
MIHIR MITHANI
Deploy Streamlit app
0d1528e
Raw History Blame Contribute Delete
2.31 kB
[
{
"id": "guard_001",
"text": "",
"expected_category": "invalid",
"description": "Empty input"
},
{
"id": "guard_002",
"text": "!!!",
"expected_category": "invalid",
"description": "Only punctuation"
},
{
"id": "guard_003",
"text": "Hello",
"expected_category": "off_topic",
"description": "Not a math problem"
},
{
"id": "guard_004",
"text": "Tell me a joke about math",
"expected_category": "off_topic",
"description": "Request for joke"
},
{
"id": "guard_005",
"text": "Write a poem about triangles",
"expected_category": "off_topic",
"description": "Request for poem"
},
{
"id": "guard_006",
"text": "What is 2+2?",
"expected_category": "incomplete",
"description": "Too short, not a word problem"
},
{
"id": "guard_007",
"text": "Solve for x",
"expected_category": "incomplete",
"description": "Missing equation"
},
{
"id": "guard_008",
"text": "Translate this to French: The cat is on the table",
"expected_category": "off_topic",
"description": "Translation request"
},
{
"id": "guard_009",
"text": "Summarize the plot of Hamlet",
"expected_category": "off_topic",
"description": "Summarization request"
},
{
"id": "guard_010",
"text": "A box has some apples. How many?",
"expected_category": "incomplete",
"description": "Missing numbers"
},
{
"id": "guard_011",
"text": "Sarah has 15 apples. She gives 4 to her friend. How many left?",
"expected_category": "math_word_problem",
"description": "Valid simple problem"
},
{
"id": "guard_012",
"text": "Write Python code to solve quadratic equations",
"expected_category": "off_topic",
"description": "Code request"
},
{
"id": "guard_013",
"text": "Explain the meaning of life",
"expected_category": "off_topic",
"description": "Philosophical question"
},
{
"id": "guard_014",
"text": " ",
"expected_category": "invalid",
"description": "Whitespace only"
},
{
"id": "guard_015",
"text": "If a train leaves Chicago at 60mph and another leaves NY at 80mph when do they meet?",
"expected_category": "math_word_problem",
"description": "Valid rate problem"
}
]