Unified Memory

Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Ask my documents

rag-019

medium answerablesingle_doc

The prompt

Document excerpts provided from: afterschool_program, school_calendar, sitter_notes

What number do I call to report Maya absent from school?
What we expected
{
  "answerable": true,
  "answer": "The attendance line, (541) 555-0251 (call by 9:00 a.m.).",
  "key_facts": [
    [
      "555-0251"
    ]
  ],
  "citations": [
    "school_calendar"
  ],
  "contra": [],
  "evidence": [
    "Attendance Line at (541) 555-0251"
  ],
  "gold_in_context": true
}

What each model answered

MiMo-V2.6 Distill 9B

✓ pass · score 100 · 3.95 s

You call the Attendance Line at (541) 555-0251, or use the ParentLink app by 9:00 a.m. [doc:school_calendar]
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar"
  ],
  "citation_recall": 1,
  "extra_citations": [],
  "citation_score": 1
}

Ornith-1.5 9B

✓ pass · score 100 · 3.16 s

You call the Attendance Line at (541) 555-0251. [doc:school_calendar]
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar"
  ],
  "citation_recall": 1,
  "extra_citations": [],
  "citation_score": 1
}

Granite 4.2 8B

✓ pass · score 100 · 5.02 s

[doc:school_calendar] states: "Please call the Attendance Line at (541) 555-0251 or use the ParentLink app by 9:00 a.m. on any day your child will be absent or late." This is the number to call to report Maya's absence.
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar"
  ],
  "citation_recall": 1,
  "extra_citations": [],
  "citation_score": 1
}

Spark-X2.5 4B

✓ pass · score 100 · 2.41 s

To report Maya absent from school, you should call the Attendance Line at (541) 555-0251 [doc:school_calendar].
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar"
  ],
  "citation_recall": 1,
  "extra_citations": [],
  "citation_score": 1
}

Mellum2.1 12B-A2.5B

✓ pass · score 100 · 1.81 s

[doc:school_calendar] says: "Attendance Line: (541) 555-0251". That is the number to report an absence. So answer: You call (541) 555-0251. Cite [doc:school_calendar].
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar"
  ],
  "citation_recall": 1,
  "extra_citations": [],
  "citation_score": 1
}

Ling-3.0-tiny (7.9B-A1.3B)

✓ pass · score 96 · 1.26 s

The Attendance Line is (541) 555-0251, as stated in [doc:school_calendar]. [doc:sitter_notes] confirms Maya's allergy but does not provide a separate reporting number.

I can't find that in your documents.
Scoring detail
{
  "answerable": true,
  "refusal": true,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar",
    "sitter_notes"
  ],
  "citation_recall": 1,
  "extra_citations": [
    "sitter_notes"
  ],
  "citation_score": 0.85
}

MiniCPM5 2B

✓ pass · score 96 · 1.43 s

The Attendance Line number is (541) 555-0251 [doc:school_calendar]. If Maya is absent, you must call this number by 9:00 a.m. [doc:school_calendar]. The after-school program notes also reference the same Attendance Line [doc:afterschool_program].
Scoring detail
{
  "answerable": true,
  "refusal": false,
  "facts_matched": [
    true
  ],
  "contra_hit": [],
  "coverage": 1,
  "cited": [
    "school_calendar",
    "afterschool_program"
  ],
  "citation_recall": 1,
  "extra_citations": [
    "afterschool_program"
  ],
  "citation_score": 0.85
}

K2-Horizon 0.9B

✗ fail · score 0 · 0.55 s

I can't find that in your documents.
Scoring detail
{
  "answerable": true,
  "refusal": true,
  "facts_matched": [
    false
  ],
  "contra_hit": [],
  "coverage": 0,
  "cited": [],
  "citation_recall": 0,
  "extra_citations": [],
  "citation_score": 0
}

← rag-018All Ask my documents itemsrag-020 →