Analysis III · Exercise research profile

Centroid, parametric integrability, Gauss and Stokes theorems (A3-25-P4-E2)

Source record, frozen scoring evidence, configuration-level measurements and retained responses for one benchmark exercise.

01

Observed panel summary

Exercise overview

Descriptive results across 21 model-mode configuration means. They do not establish general model superiority.

Published recordSource statement · frozen rubric · 23 retained responses · current judgmentsThe separate official reference solution is not reproduced.
Mean rubric score85.1Median 93.5 · n=16
Configuration spreadSD 19.435.3–100.0
Rank by lowest observed mean#12/60Overall · #9/33 within Analysis III
Mean latency177.8sEnd-to-end per response
Mean cost$0.121Snapshot estimate
Mean tokens16,439Analytical tokens per response

Configuration means

Observed score distribution

Median 93.5 / 100
AnthropicOpenAIGoogleDeepSeekxAIMistral
  1. Claude Fable 5SD not available · single run
    100.0
  2. Claude Opus 5.5SD not available · single run
    100.0
  3. Gemini 3.1 Pro PreviewSD not available · single run
    100.0
  4. Gemini 3.5 FlashSD 0.0 · 3 retained runs
    100.0
  5. GPT-5.6 SolSD not available · single run
    100.0
  6. GPT-6.1 SolSD not available · single run
    100.0
  7. Claude Sonnet 5.5SD not available · single run
    99.3
  8. DeepSeek V4 ProSD not available · single run
    98.8
  9. GPT-6 LunaSD not available · single run
    98.8
  10. DeepSeek V4.1 FlashSD not available · single run
    96.5
  11. DeepSeek V4 FlashSD not available · single run
    93.5
  12. Grok 4.3SD not available · single run
    92.0
  13. Grok 4.5SD not available · single run
    91.0
  14. OpenAI o3SD not available · single run
    82.0
  15. Mistral Medium 3.5SD not available · single run
    81.0
  16. Mistral Small 4SD not available · single run
    80.0
  17. Gemma 4 31B ITSD not available · single run
    76.8
  18. DeepSeek V4 Flash (no thinking)SD not available · single run
    68.0
  19. OpenAI GPT-4.1SD not available · single run
    48.8
  20. Claude Sonnet 4.5SD not available · single run
    46.0
  21. Claude Haiku 4.5SD not available · single run
    35.3

Whiskers show mean ± one within-configuration response SD where three retained runs are available. Single-run cells have no within-configuration variability statistic.

02

Task before results

Source record and frozen rubric

Read the evaluated task before interpreting configuration outcomes. Original Italian records remain authoritative.

Evaluated source record

Exercise statement

A3-25-P4-E2 · Analysis III
The English version is a faithful editorial translation. The Italian source remains authoritative.

ESAME: Analisi 3 DATA: 03/07/2025 ESERCIZIO: 2

TESTO: 2. Nel piano cartesiano sia A = A1∪A2 con A1 ={(x,y ) : 0 ≤x≤a, 0≤y ≤a−x}e A2 ={(x,y ) :x≤0, y≥0, x2 +y2≤a2}(ovea> 0). (a) Disegnare A e calcolarne il baricentro geometrico. (b) Calcolare (se converge) ∫ A x√ydxdy ; più in generale, dire per qualiα∈R converge ∫ Axyαdxdy . Si disegni ora nello spazio cartesiano il cono C avente come base A nel piano orizzontale (x,y ) e come vertice il punto (0, 0,a ) sull’assez. (c) Verificare il teorema di Gauss per C e per il campo F = (x +y, 0, 0). (d) Verificare la formula di Kelvin-Stokes per F e la componente piana obliqua S di ∂C.

Source transcription errata1 recorded normalization · original preserved+

These notes document evident transcription or OCR artifacts. They do not overwrite the authoritative source record above.

As transcribedA1∪A2; x2+y2≤a2; x√ydxdy; yα; sull’assezEditorial normalizationA₁∪A₂; x²+y²≤a²; x√y dx dy; y^α; sull’asse zLost subscripts, superscripts and mathematical spacing.
Original source textExact Italian record · no editorial changes+
ESAME: Analisi 3
DATA: 03/07/2025
ESERCIZIO: 2

TESTO:
2. Nel piano cartesiano sia A = A1∪A2 con A1 ={(x,y ) : 0 ≤x≤a, 0≤y ≤a−x}e
A2 ={(x,y ) :x≤0, y≥0, x2 +y2≤a2}(ovea> 0).
(a) Disegnare A e calcolarne il baricentro geometrico.
(b) Calcolare (se converge)
∫
A
x√ydxdy ; più in generale, dire per qualiα∈R converge
∫
Axyαdxdy .
Si disegni ora nello spazio cartesiano il cono C avente come base A nel piano orizzontale (x,y ) e
come vertice il punto (0, 0,a ) sull’assez.
(c) Verificare il teorema di Gauss per C e per il campo F = (x +y, 0, 0).
(d) Verificare la formula di Kelvin-Stokes per F e la componente piana obliqua S di ∂C.
TopicCentroid, parametric integrability, Gauss and Stokes theorems
Syllabus areaMultiple integrals and integral theorems
Task typemixed
Atomic criteria30
Subproblem a · 20.0%

Disegno e baricentro.

7 atomic criteria · T1 · T2 · T3 · T4 · T5 · T6 · T7

Inspect scoring criteria +
  • Descrive A1 e A2.

    T1 · 5 pts · Criterion (a_T1_01)
  • Imposta area e momenti.

    T2 · 3 pts · Criterion (a_T2_01)
  • Calcola area e momenti.

    T3 · 4 pts · Criterion (a_T3_01)
  • Dà G.

    T4 · 3 pts · Criterion (a_T4_01)
  • Rispetta domini.

    T5 · 2 pts · Criterion (a_T5_01)
  • Controlla scala/posizione.

    T6 · 2 pts · Criterion (a_T6_01)
  • Usa definizione di baricentro.

    T7 · 1 pts · Criterion (a_T7_01)
Subproblem b · 24.0%

Integrale particolare e soglia α.

7 atomic criteria · T1 · T2 · T3 · T4 · T5 · T6 · T7

Inspect scoring criteria +
  • Localizza le singolarita della famiglia generale e tratta coerentemente il particolare.

    T1 · 5 pts · Criterion (b_T1_01)
  • Imposta gli integrali iterati con una delle due letture ammesse.

    T2 · 4 pts · Criterion (b_T2_01)
  • Trova separatamente la soglia generale e il valore particolare ammesso.

    T3 · 5 pts · Criterion (b_T3_01)
  • Dà conclusioni complete senza richiedere entrambe le letture del particolare.

    T4 · 3 pts · Criterion (b_T4_01)
  • Gestisce l'improprieta della famiglia generale separatamente dall'ambiguita del particolare.

    T5 · 3 pts · Criterion (b_T5_01)
  • Controlla segno e scala rispetto alla lettura scelta.

    T6 · 2 pts · Criterion (b_T6_01)
  • Applica coerentemente i criteri di integrabilita senza mescolare le due interpretazioni.

    T7 · 2 pts · Criterion (b_T7_01)
Subproblem c · 33.0%

Verificare Gauss.

8 atomic criteria · T1 · T2 · T3 · T4 · T5 · T6 · T7 · T8

Inspect scoring criteria +
  • Identifica cono e frontiera.

    T1 · 6 pts · Criterion (c_T1_01)
  • Imposta volume e flussi.

    T2 · 5 pts · Criterion (c_T2_01)
  • Calcola volume e flussi.

    T3 · 7 pts · Criterion (c_T3_01)
  • Conclude Gauss.

    T4 · 5 pts · Criterion (c_T4_01)
  • Usa domini/orientazioni.

    T5 · 3 pts · Criterion (c_T5_01)
  • Controlla segni/scala.

    T6 · 2 pts · Criterion (c_T6_01)
  • Applica Gauss rigorosamente.

    T7 · 2 pts · Criterion (c_T7_01)
  • Verifica autonomamente Gauss.

    T8 · 3 pts · Criterion (c_T8_01)
Subproblem d · 23.0%

Verificare Kelvin-Stokes.

8 atomic criteria · T1 · T2 · T3 · T4 · T5 · T6 · T7 · T8

Inspect scoring criteria +
  • Identifica S,bordo,versi.

    T1 · 4 pts · Criterion (d_T1_01)
  • Imposta i due membri.

    T2 · 3 pts · Criterion (d_T2_01)
  • Calcola entrambi.

    T3 · 4 pts · Criterion (d_T3_01)
  • Conclude Stokes.

    T4 · 4 pts · Criterion (d_T4_01)
  • Mantiene orientazioni.

    T5 · 2 pts · Criterion (d_T5_01)
  • Controlla segno/scala.

    T6 · 2 pts · Criterion (d_T6_01)
  • Applica Stokes rigorosamente.

    T7 · 2 pts · Criterion (d_T7_01)
  • Verifica autonomamente Stokes.

    T8 · 2 pts · Criterion (d_T8_01)
Complete frozen rubricOriginal Italian record · exact JSON · SHA-256 4ffefc2fa1ba…+
{
  "rubric_version": "1.1-adjudication-2026-08-10",
  "judge_spec_version": "2.0",
  "item_id": "analisi_3_2025_quarto_appello_esercizio_2",
  "syllabus_macroarea": "Integrali multipli e teoremi integrali",
  "topic": "Baricentro, integrabilità parametrica, Gauss e Stokes",
  "task_type": "mixed",
  "created_without_model_answers": true,
  "frozen": true,
  "reference_status": "corrected_adjudication_current_authoritative",
  "reference_notes": "Decisione REFERENCE_ERROR: il testo del benchmark e ambiguo tra x/sqrt(y) e x*sqrt(y). Sono ammesse entrambe le letture: -4a^(5/2)/15 per x/sqrt(y) e -4a^(7/2)/35 per x*sqrt(y), con credito pieno se procedimento e risultato sono coerenti con una delle due. Non sono ammessi calcoli misti. La soglia generale per x*y^alpha resta alpha>-1 ed e valutata separatamente. Tutte le altre parti della reference e della rubric restano invariate.",
  "subproblems": [
    {
      "id": "a",
      "weight": 0.2,
      "request": "Disegno e baricentro.",
      "expected_results": [
        "area,momenti,G gold"
      ],
      "required_conditions": [
        "a>0."
      ],
      "accepted_equivalences": [
        "Forme equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodi rigorosi compatibili col gold."
      ]
    },
    {
      "id": "b",
      "weight": 0.24,
      "request": "Integrale particolare e soglia α.",
      "expected_results": [
        "Soglia generale, valutata separatamente: alpha>-1",
        "Se il particolare e letto x/sqrt(y): -4a^(5/2)/15",
        "Se il particolare e letto x*sqrt(y): -4a^(7/2)/35"
      ],
      "required_conditions": [
        "Separare A1 e A2.",
        "Usare coerentemente una sola interpretazione del particolare.",
        "Non mescolare le due interpretazioni."
      ],
      "accepted_equivalences": [
        "Qualsiasi forma equivalente di uno dei due valori ammessi, con scala a^(5/2) per x/sqrt(y) oppure a^(7/2) per x*sqrt(y)."
      ],
      "accepted_alternative_methods": [
        "Metodi rigorosi coerenti con una delle due letture; non e richiesto calcolare entrambe."
      ]
    },
    {
      "id": "c",
      "weight": 0.33,
      "request": "Verificare Gauss.",
      "expected_results": [
        "volume e flussi gold",
        "uguaglianza indipendente"
      ],
      "required_conditions": [
        "Normali uscenti."
      ],
      "accepted_equivalences": [
        "Forme equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodi rigorosi compatibili col gold."
      ]
    },
    {
      "id": "d",
      "weight": 0.23,
      "request": "Verificare Kelvin-Stokes.",
      "expected_results": [
        "rotore (0,0,-1)",
        "entrambi -a²/2"
      ],
      "required_conditions": [
        "Orientazioni coerenti."
      ],
      "accepted_equivalences": [
        "Forme equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodi rigorosi compatibili col gold."
      ]
    }
  ],
  "atomic_criteria": [
    {
      "criterion_id": "a_T1_01",
      "subproblem_id": "a",
      "primary_dimension": "T1",
      "max_points": 5,
      "description": "Descrive A1 e A2.",
      "full_credit_evidence": [
        "Triangolo e quarto di disco."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T2_01",
      "subproblem_id": "a",
      "primary_dimension": "T2",
      "max_points": 3,
      "description": "Imposta area e momenti.",
      "full_credit_evidence": [
        "Scomposizione corretta."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T3_01",
      "subproblem_id": "a",
      "primary_dimension": "T3",
      "max_points": 4,
      "description": "Calcola area e momenti.",
      "full_credit_evidence": [
        "Valori gold."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T4_01",
      "subproblem_id": "a",
      "primary_dimension": "T4",
      "max_points": 3,
      "description": "Dà G.",
      "full_credit_evidence": [
        "Baricentro gold."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T5_01",
      "subproblem_id": "a",
      "primary_dimension": "T5",
      "max_points": 2,
      "description": "Rispetta domini.",
      "full_credit_evidence": [
        "Limiti cartesiani/polari."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T6_01",
      "subproblem_id": "a",
      "primary_dimension": "T6",
      "max_points": 2,
      "description": "Controlla scala/posizione.",
      "full_credit_evidence": [
        "Quantità dimensionalmente coerenti."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "a_T7_01",
      "subproblem_id": "a",
      "primary_dimension": "T7",
      "max_points": 1,
      "description": "Usa definizione di baricentro.",
      "full_credit_evidence": [
        "Momenti/area."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T1_01",
      "subproblem_id": "b",
      "primary_dimension": "T1",
      "max_points": 5,
      "description": "Localizza le singolarita della famiglia generale e tratta coerentemente il particolare.",
      "full_credit_evidence": [
        "Per x*y^alpha analizza y=0 e rho=0; per il particolare segue coerentemente alpha=-1/2 oppure alpha=1/2."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T2_01",
      "subproblem_id": "b",
      "primary_dimension": "T2",
      "max_points": 4,
      "description": "Imposta gli integrali iterati con una delle due letture ammesse.",
      "full_credit_evidence": [
        "Cartesiane e/o polari coerenti: esponente -1/2 per x/sqrt(y), oppure +1/2 per x*sqrt(y)."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T3_01",
      "subproblem_id": "b",
      "primary_dimension": "T3",
      "max_points": 5,
      "description": "Trova separatamente la soglia generale e il valore particolare ammesso.",
      "full_credit_evidence": [
        "alpha>-1 e, coerentemente con la lettura scelta, -4a^(5/2)/15 oppure -4a^(7/2)/35."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T4_01",
      "subproblem_id": "b",
      "primary_dimension": "T4",
      "max_points": 3,
      "description": "Dà conclusioni complete senza richiedere entrambe le letture del particolare.",
      "full_credit_evidence": [
        "Soglia iff alpha>-1 e valore coerente con una delle due letture; non e necessario presentarle entrambe."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T5_01",
      "subproblem_id": "b",
      "primary_dimension": "T5",
      "max_points": 3,
      "description": "Gestisce l'improprieta della famiglia generale separatamente dall'ambiguita del particolare.",
      "full_credit_evidence": [
        "Tonelli/Fubini o confronti corretti per concludere alpha>-1; nessuna penalita per la lettura x*sqrt(y) del particolare."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T6_01",
      "subproblem_id": "b",
      "primary_dimension": "T6",
      "max_points": 2,
      "description": "Controlla segno e scala rispetto alla lettura scelta.",
      "full_credit_evidence": [
        "Valore negativo con scala a^(5/2) per x/sqrt(y), oppure a^(7/2) per x*sqrt(y)."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T7_01",
      "subproblem_id": "b",
      "primary_dimension": "T7",
      "max_points": 2,
      "description": "Applica coerentemente i criteri di integrabilita senza mescolare le due interpretazioni.",
      "full_credit_evidence": [
        "Nessuna cancellazione impropria e nessun passaggio che combini esponenti o risultati delle due letture."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti coerenti con una delle due letture ammesse."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso coerente con x/sqrt(y) oppure x*sqrt(y), senza mescolare le letture."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T1_01",
      "subproblem_id": "c",
      "primary_dimension": "T1",
      "max_points": 6,
      "description": "Identifica cono e frontiera.",
      "full_credit_evidence": [
        "A,B,T,S."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T2_01",
      "subproblem_id": "c",
      "primary_dimension": "T2",
      "max_points": 5,
      "description": "Imposta volume e flussi.",
      "full_credit_evidence": [
        "Parametrizzazioni/normali."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T3_01",
      "subproblem_id": "c",
      "primary_dimension": "T3",
      "max_points": 7,
      "description": "Calcola volume e flussi.",
      "full_credit_evidence": [
        "Tutti i valori gold."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T4_01",
      "subproblem_id": "c",
      "primary_dimension": "T4",
      "max_points": 5,
      "description": "Conclude Gauss.",
      "full_credit_evidence": [
        "Somma=Vol."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T5_01",
      "subproblem_id": "c",
      "primary_dimension": "T5",
      "max_points": 3,
      "description": "Usa domini/orientazioni.",
      "full_credit_evidence": [
        "θ,z e A1 corretti."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T6_01",
      "subproblem_id": "c",
      "primary_dimension": "T6",
      "max_points": 2,
      "description": "Controlla segni/scala.",
      "full_credit_evidence": [
        "Contributi ~a³."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T7_01",
      "subproblem_id": "c",
      "primary_dimension": "T7",
      "max_points": 2,
      "description": "Applica Gauss rigorosamente.",
      "full_credit_evidence": [
        "Ipotesi e frontiera."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T8_01",
      "subproblem_id": "c",
      "primary_dimension": "T8",
      "max_points": 3,
      "description": "Verifica autonomamente Gauss.",
      "full_credit_evidence": [
        "Due membri indipendenti."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T1_01",
      "subproblem_id": "d",
      "primary_dimension": "T1",
      "max_points": 4,
      "description": "Identifica S,bordo,versi.",
      "full_credit_evidence": [
        "Tre tratti coerenti."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T2_01",
      "subproblem_id": "d",
      "primary_dimension": "T2",
      "max_points": 3,
      "description": "Imposta i due membri.",
      "full_credit_evidence": [
        "Flusso rotore e circolazione."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T3_01",
      "subproblem_id": "d",
      "primary_dimension": "T3",
      "max_points": 4,
      "description": "Calcola entrambi.",
      "full_credit_evidence": [
        "-a²/2."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T4_01",
      "subproblem_id": "d",
      "primary_dimension": "T4",
      "max_points": 4,
      "description": "Conclude Stokes.",
      "full_credit_evidence": [
        "Uguaglianza esplicita."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T5_01",
      "subproblem_id": "d",
      "primary_dimension": "T5",
      "max_points": 2,
      "description": "Mantiene orientazioni.",
      "full_credit_evidence": [
        "Normale-bordo compatibili."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T6_01",
      "subproblem_id": "d",
      "primary_dimension": "T6",
      "max_points": 2,
      "description": "Controlla segno/scala.",
      "full_credit_evidence": [
        "Entrambi ~a²."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T7_01",
      "subproblem_id": "d",
      "primary_dimension": "T7",
      "max_points": 2,
      "description": "Applica Stokes rigorosamente.",
      "full_credit_evidence": [
        "Ipotesi corrette."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "d_T8_01",
      "subproblem_id": "d",
      "primary_dimension": "T8",
      "max_points": 2,
      "description": "Verifica autonomamente Stokes.",
      "full_credit_evidence": [
        "Calcoli indipendenti."
      ],
      "partial_credit_rules": [
        {
          "credit_fraction": 0.75,
          "anchor": "Corretto salvo un difetto locale."
        },
        {
          "credit_fraction": 0.5,
          "anchor": "Copertura sostanziale ma incompleta."
        },
        {
          "credit_fraction": 0.25,
          "anchor": "Solo impostazione o elemento utile."
        }
      ],
      "zero_credit_conditions": [
        "Assente o incompatibile col gold."
      ],
      "accepted_equivalences": [
        "Forme matematicamente equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodo alternativo rigoroso compatibile col gold."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    }
  ],
  "dimension_applicability": {
    "T1": "applicable",
    "T2": "applicable",
    "T3": "applicable",
    "T4": "applicable",
    "T5": "applicable",
    "T6": "applicable",
    "T7": "applicable",
    "T8": "applicable"
  },
  "dimension_applicability_notes": {
    "T8": "Applicabile per le verifiche esplicitamente richieste."
  },
  "dimension_point_budget": {
    "T1": 20,
    "T2": 15,
    "T3": 20,
    "T4": 15,
    "T5": 10,
    "T6": 8,
    "T7": 7,
    "T8": 5
  },
  "applicable_max_score": 100,
  "normalization_rule": "Punteggio grezzo su 100.",
  "not_applicable_dimensions": [],
  "pertinent_checks": [
    "Area e momenti.",
    "Singolarità A1/A2.",
    "Normali di Gauss.",
    "Orientazione Stokes."
  ],
  "known_error_severity": [
    {
      "error": "Soglia errata",
      "severity": "major"
    },
    {
      "error": "Faccia omessa",
      "severity": "major"
    },
    {
      "error": "Verifica solo dichiarata",
      "severity": "major"
    },
    {
      "error": "Orientazione incoerente",
      "severity": "major"
    }
  ],
  "rubric_specific_caps": [],
  "human_review_triggers": [
    "Parametrizzazione/orientazione alternativa incerta.",
    "Risposta influenzata dall'ambiguità di trascrizione x√y rispetto al gold x/√y."
  ],
  "human_approval": {
    "approved": true,
    "reviewer": "benchmark_owner_via_codex_task",
    "date": "2026-08-10",
    "approval_source": "Explicit instruction 'procedi, quelle storiche lasciale cosi' in the current Codex task.",
    "decision_id": "2026-08-10_quartoapp_esercizio_2_dual_interpretation_v1"
  }
}
03

Per-exercise ranking

Configurations

Choose one of four stable ranking views. Retry count remains available as a secondary table sort.

Rank configurations by

Score ranks high to low. Speed, cost and tokens rank low to high. Retry sorting remains available from the table header.

Results for all 21 model-mode configurations
ConfigurationSelected rankMean score / replicasLatencyCostTokens
Claude Fable 5Anthropic · official model page ↗#1by score100.0SD not available · single run112.6s$0.54511,6710.00
Claude Opus 5.5Anthropic · official model page ↗#1by score100.0SD not available · single run93.1s$0.19810,6510.00
Gemini 3.1 Pro PreviewGoogle · official model page ↗#1by score100.0SD not available · single run103.6s$0.16414,1780.00
Gemini 3.5 FlashGoogle · official model page ↗#1by score100.0SD 0.0 · 3 retained runs616.1s$0.18220,6960.00
GPT-5.6 SolOpenAI · official model page ↗#1by score100.0SD not available · single run323.3s$0.58820,0980.00
GPT-6.1 SolOpenAI · official model page ↗#1by score100.0SD not available · single run191.7s$0.10611,0810.00
Claude Sonnet 5.5Anthropic · official model page ↗#7by score99.3SD not available · single run46.7s$0.0717,8380.00
DeepSeek V4 ProDeepSeek · official model page ↗#8by score98.8SD not available · single run450.5s$0.02731,4320.00
GPT-6 LunaOpenAI · official model page ↗#8by score98.8SD not available · single run124.8s$0.007515,5150.00
DeepSeek V4.1 FlashDeepSeek · official model page ↗#10by score96.5SD not available · single run101.3s$0.02823,9610.00
DeepSeek V4 FlashDeepSeek · official model page ↗#11by score93.5SD not available · single run324.8s$0.009734,9190.00
Grok 4.3xAI · official model page ↗#12by score92.0SD not available · single run145.3s$0.04317,7160.00
Grok 4.5xAI · official model page ↗#13by score91.0SD not available · single run321.5s$0.11820,2840.00
OpenAI o3OpenAI · official model page ↗#14by score82.0SD not available · single run123.7s$0.11915,3930.00
Mistral Medium 3.5Mistral · official model page ↗#15by score81.0SD not available · single run158.0s$0.17623,9670.00
Mistral Small 4Mistral · official model page ↗#16by score80.0SD not available · single run159.2s$0.01830,9090.00
Gemma 4 31B ITGoogle · official model page ↗#17by score76.8SD not available · single run104.0s$0.00004,3160.00
DeepSeek V4 Flash (no thinking)DeepSeek · official model page ↗#18by score68.0SD not available · single run105.3s$0.003813,9320.00
OpenAI GPT-4.1OpenAI · official model page ↗#19by score48.8SD not available · single run57.4s$0.0587,6990.00
Claude Sonnet 4.5Anthropic · official model page ↗#20by score46.0SD not available · single run43.6s$0.0483,7820.00
Claude Haiku 4.5Anthropic · official model page ↗#21by score35.3SD not available · single run27.9s$0.0235,1870.00
#1 by score

Claude Fable 5

Anthropic
100.0mean score

SD not available · single run

Latency
112.6s
Cost
$0.545
Tokens
11,671
Retries
0.00
Official model page ↗
#1 by score

Claude Opus 5.5

Anthropic
100.0mean score

SD not available · single run

Latency
93.1s
Cost
$0.198
Tokens
10,651
Retries
0.00
Official model page ↗
#1 by score

Gemini 3.1 Pro Preview

Google
100.0mean score

SD not available · single run

Latency
103.6s
Cost
$0.164
Tokens
14,178
Retries
0.00
Official model page ↗
#1 by score

Gemini 3.5 Flash

Google
100.0mean score

SD 0.0 · 3 retained runs

Latency
616.1s
Cost
$0.182
Tokens
20,696
Retries
0.00
Official model page ↗
#1 by score

GPT-5.6 Sol

OpenAI
100.0mean score

SD not available · single run

Latency
323.3s
Cost
$0.588
Tokens
20,098
Retries
0.00
Official model page ↗
#1 by score

GPT-6.1 Sol

OpenAI
100.0mean score

SD not available · single run

Latency
191.7s
Cost
$0.106
Tokens
11,081
Retries
0.00
Official model page ↗
#7 by score

Claude Sonnet 5.5

Anthropic
99.3mean score

SD not available · single run

Latency
46.7s
Cost
$0.071
Tokens
7,838
Retries
0.00
Official model page ↗
#8 by score

DeepSeek V4 Pro

DeepSeek
98.8mean score

SD not available · single run

Latency
450.5s
Cost
$0.027
Tokens
31,432
Retries
0.00
Official model page ↗
#8 by score

GPT-6 Luna

OpenAI
98.8mean score

SD not available · single run

Latency
124.8s
Cost
$0.0075
Tokens
15,515
Retries
0.00
Official model page ↗
#10 by score

DeepSeek V4.1 Flash

DeepSeek
96.5mean score

SD not available · single run

Latency
101.3s
Cost
$0.028
Tokens
23,961
Retries
0.00
Official model page ↗
#11 by score

DeepSeek V4 Flash

DeepSeek
93.5mean score

SD not available · single run

Latency
324.8s
Cost
$0.0097
Tokens
34,919
Retries
0.00
Official model page ↗
#12 by score

Grok 4.3

xAI
92.0mean score

SD not available · single run

Latency
145.3s
Cost
$0.043
Tokens
17,716
Retries
0.00
Official model page ↗
#13 by score

Grok 4.5

xAI
91.0mean score

SD not available · single run

Latency
321.5s
Cost
$0.118
Tokens
20,284
Retries
0.00
Official model page ↗
#14 by score

OpenAI o3

OpenAI
82.0mean score

SD not available · single run

Latency
123.7s
Cost
$0.119
Tokens
15,393
Retries
0.00
Official model page ↗
#15 by score

Mistral Medium 3.5

Mistral
81.0mean score

SD not available · single run

Latency
158.0s
Cost
$0.176
Tokens
23,967
Retries
0.00
Official model page ↗
#16 by score

Mistral Small 4

Mistral
80.0mean score

SD not available · single run

Latency
159.2s
Cost
$0.018
Tokens
30,909
Retries
0.00
Official model page ↗
#17 by score

Gemma 4 31B IT

Google
76.8mean score

SD not available · single run

Latency
104.0s
Cost
$0.0000
Tokens
4,316
Retries
0.00
Official model page ↗
#18 by score

DeepSeek V4 Flash (no thinking)

DeepSeek
68.0mean score

SD not available · single run

Latency
105.3s
Cost
$0.0038
Tokens
13,932
Retries
0.00
Official model page ↗
#19 by score

OpenAI GPT-4.1

OpenAI
48.8mean score

SD not available · single run

Latency
57.4s
Cost
$0.058
Tokens
7,699
Retries
0.00
Official model page ↗
#20 by score

Claude Sonnet 4.5

Anthropic
46.0mean score

SD not available · single run

Latency
43.6s
Cost
$0.048
Tokens
3,782
Retries
0.00
Official model page ↗
#21 by score

Claude Haiku 4.5

Anthropic
35.3mean score

SD not available · single run

Latency
27.9s
Cost
$0.023
Tokens
5,187
Retries
0.00
Official model page ↗
04

Criterion-level benchmark record

Scoring evidence

Dimension and criterion values are descriptive results for the 21 evaluated configuration means. N/A remains distinct from zero.

Attainment scale0%50%100%N/A = not applicable
T184.2%Understanding / modelling
T288.2%Method selection
T385.9%Development
T484.2%Final results
T582.9%Domains / assumptions
T692.0%Consistency
T784.3%Rigor
T874.0%Independent verification
T1 through T8 attainment for each model-mode configuration
ConfigurationT120 ptsT215 ptsT320 ptsT415 ptsT510 ptsT68 ptsT77 ptsT85 pts
Claude Fable 5100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
Claude Opus 5.5100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
GPT-5.6 Sol100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
GPT-6.1 Sol100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
Gemini 3.1 Pro Preview100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
Gemini 3.5 Flash100.0%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
Claude Sonnet 5.5100.0%100.0%100.0%100.0%92.5%100.0%100.0%100.0%
DeepSeek V4 Pro93.8%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
GPT-6 Luna93.8%100.0%100.0%100.0%100.0%100.0%100.0%100.0%
DeepSeek V4.1 Flash100.0%100.0%100.0%90.0%100.0%100.0%71.4%100.0%
DeepSeek V4 Flash93.8%100.0%93.8%90.0%85.0%100.0%85.7%100.0%
Grok 4.380.0%100.0%100.0%73.3%100.0%100.0%100.0%100.0%
Grok 4.5100.0%100.0%87.5%90.0%70.0%100.0%71.4%100.0%
OpenAI o385.0%75.0%82.5%100.0%77.5%87.5%78.6%40.0%
Mistral Medium 3.585.0%83.3%82.5%83.3%70.0%100.0%78.6%40.0%
Mistral Small 468.8%73.3%82.5%100.0%72.5%93.8%78.6%70.0%
Gemma 4 31B IT65.0%83.3%82.5%73.3%77.5%100.0%85.7%40.0%
DeepSeek V4 Flash (no thinking)67.5%61.7%72.5%70.0%70.0%62.5%78.6%55.0%
OpenAI GPT-4.152.5%65.0%47.5%30.0%50.0%75.0%50.0%0.0%
Claude Sonnet 4.550.0%65.0%41.3%45.0%42.5%50.0%42.9%0.0%
Claude Haiku 4.532.5%45.0%31.3%23.3%32.5%62.5%50.0%10.0%
05

Load on request

Retained responses and judgments

23 responses across 21 model-mode configurations. Long records load only when opened.

Interpretation boundary

Answers are reproduced as submitted. Each is paired with its authority-resolved judgment for this snapshot; this is not a general evaluation of the underlying model.

Compare two configurations All retained replicas

Select two configurations and load all their retained replicas.

Select two configurations and load all their retained replicas.

#1Claude Fable 5Anthropic · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1Claude Opus 5.5Anthropic · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1GPT-5.6 SolOpenAI · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1GPT-6.1 SolOpenAI · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1Gemini 3.1 Pro PreviewGoogle · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1Gemini 3.5 FlashGoogle · 3 retained runs · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#7Claude Sonnet 5.5Anthropic · 1 retained run · load on request99.3configuration mean+
Configuration recordOfficial model page ↗
#8DeepSeek V4 ProDeepSeek · 1 retained run · load on request98.8configuration mean+
Configuration recordOfficial model page ↗
#8GPT-6 LunaOpenAI · 1 retained run · load on request98.8configuration mean+
Configuration recordOfficial model page ↗
#10DeepSeek V4.1 FlashDeepSeek · 1 retained run · load on request96.5configuration mean+
Configuration recordOfficial model page ↗
#11DeepSeek V4 FlashDeepSeek · 1 retained run · load on request93.5configuration mean+
Configuration recordOfficial model page ↗
#12Grok 4.3xAI · 1 retained run · load on request92.0configuration mean+
Configuration recordOfficial model page ↗
#13Grok 4.5xAI · 1 retained run · load on request91.0configuration mean+
Configuration recordOfficial model page ↗
#14OpenAI o3OpenAI · 1 retained run · load on request82.0configuration mean+
Configuration recordOfficial model page ↗
#15Mistral Medium 3.5Mistral · 1 retained run · load on request81.0configuration mean+
Configuration recordOfficial model page ↗
#16Mistral Small 4Mistral · 1 retained run · load on request80.0configuration mean+
Configuration recordOfficial model page ↗
#17Gemma 4 31B ITGoogle · 1 retained run · load on request76.8configuration mean+
Configuration recordOfficial model page ↗
#18DeepSeek V4 Flash (no thinking)DeepSeek · 1 retained run · load on request68.0configuration mean+
Configuration recordOfficial model page ↗
#19OpenAI GPT-4.1OpenAI · 1 retained run · load on request48.8configuration mean+
Configuration recordOfficial model page ↗
#20Claude Sonnet 4.5Anthropic · 1 retained run · load on request46.0configuration mean+
Configuration recordOfficial model page ↗
#21Claude Haiku 4.5Anthropic · 1 retained run · load on request35.3configuration mean+
Configuration recordOfficial model page ↗
06

Record context

About this benchmark record

The Atlas presents the exercise, rubric, results and retained responses in one interactive record.

Exercise IDA3-25-P4-E2
SubjectAnalysis III
Exam sitting2025 · Fourth sitting
Benchmark snapshot2026-10-15
Rubric version1.1-adjudication-2026-08-10
Retained responses23
How to use this page

Use the Atlas to inspect what was tested and how each configuration performed. Most configuration–exercise cells are single-run, so displayed values are descriptive rather than uncertainty estimates.