Analisi III · Exercise research profile

Esistenza, energia e ODE lineare forzata (A3-25-P4-E3)

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 score84.0Median 83.7 · n=16
Configuration spreadSD 15.145.0–100.0
Rank by lowest observed mean#9/60Overall · #7/33 within Analisi III
Mean latency189.3sEnd-to-end per response
Mean cost$0.095Snapshot estimate
Mean tokens10,548Analytical tokens per response

Configuration means

Observed score distribution

Median 83.7 / 100
AnthropicDeepSeekOpenAIxAIGoogleMistral
  1. Claude Fable 5SD not available · single run
    100.0
  2. Claude Sonnet 5.5SD not available · single run
    100.0
  3. DeepSeek V4.1 FlashSD not available · single run
    100.0
  4. GPT-5.6 SolSD not available · single run
    100.0
  5. GPT-6.1 SolSD not available · single run
    100.0
  6. DeepSeek V4 FlashSD not available · single run
    98.7
  7. Claude Opus 5.5SD not available · single run
    97.4
  8. Grok 4.3SD not available · single run
    93.2
  9. GPT-6 LunaSD not available · single run
    92.6
  10. Gemini 3.5 FlashSD 12.7 · 3 retained runs
    85.6
  11. DeepSeek V4 ProSD not available · single run
    83.7
  12. Grok 4.5SD not available · single run
    82.9
  13. DeepSeek V4 Flash (no thinking)SD not available · single run
    82.1
  14. Gemini 3.1 Pro PreviewSD not available · single run
    81.8
  15. OpenAI o3SD not available · single run
    80.5
  16. OpenAI GPT-4.1SD not available · single run
    74.7
  17. Mistral Small 4SD not available · single run
    72.9
  18. Mistral Medium 3.5SD not available · single run
    69.5
  19. Claude Haiku 4.5SD not available · single run
    65.5
  20. Gemma 4 31B ITSD not available · single run
    58.2
  21. Claude Sonnet 4.5SD not available · single run
    45.0

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-E3 · Analisi III
The English version is a faithful editorial translation. The Italian source remains authoritative.

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

TESTO: 3. Si abbia il problema di Cauchy { y′′= 2(2y + 3α√y + 1) + (α−1)e−2t y(0) = 1, y′(0) =−4 in y(t) (ove α∈R). (a) Cosa si può affermare a priori su esistenza e unicità della soluzione del problema dato? (b) Risolvere il problema di Cauchy quando α= 1. (c) Risolvere il problema di Cauchy quando α= 0.

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 transcribed(α−1)e−2t; ove α∈REditorial normalization(α−1)e^(−2t); ove α∈ℝThe exponential superscript and number-set styling were flattened.
Original source textExact Italian record · no editorial changes+
ESAME: Analisi 3
DATA: 03/07/2025
ESERCIZIO: 3

TESTO:
3. Si abbia il problema di Cauchy
{
y′′= 2(2y + 3α√y + 1) + (α−1)e−2t
y(0) = 1, y′(0) =−4 in y(t) (ove α∈R).
(a) Cosa si può affermare a priori su esistenza e unicità della soluzione del problema dato?
(b) Risolvere il problema di Cauchy quando α= 1.
(c) Risolvere il problema di Cauchy quando α= 0.
TopicEsistenza, energia e ODE lineare forzata
Syllabus areaEquazioni differenziali del secondo ordine
Task typemixed
Atomic criteria20
Subproblem a · 28.4%

Conclusioni a priori.

6 atomic criteria · T1 · T2 · T4 · T5 · T6 · T7

Inspect scoring criteria +
  • Riduce a sistema e identifica regolarità.

    T1 · 10 pts · Criterion (a_T1_01)
  • Sceglie teoremi appropriati.

    T2 · 5 pts · Criterion (a_T2_01)
  • Dà conclusioni a priori.

    T4 · 3 pts · Criterion (a_T4_01)
  • Gestisce dominio del radicale.

    T5 · 4 pts · Criterion (a_T5_01)
  • Distingue locale/globale.

    T6 · 2 pts · Criterion (a_T6_01)
  • Applica rigorosamente i teoremi.

    T7 · 3 pts · Criterion (a_T7_01)
Subproblem b · 35.8%

Risolvere α=1.

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

Inspect scoring criteria +
  • Riconosce y''=h(y).

    T1 · 5 pts · Criterion (b_T1_01)
  • Imposta energia, separazione e cambio di ramo al raccordo.

    T2 · 5 pts · Criterion (b_T2_01)
  • Calcola entrambi i rami e verifica il secondo nell'ODE.

    T3 · 10 pts · Criterion (b_T3_01)
  • Dà la soluzione massimale globale a tratti.

    T4 · 6 pts · Criterion (b_T4_01)
  • Gestisce correttamente raccordo, regolarita e dominio.

    T5 · 3 pts · Criterion (b_T5_01)
  • Verifica dati iniziali, energia e condizioni di raccordo.

    T6 · 3 pts · Criterion (b_T6_01)
  • Giustifica rigorosamente l'unicita del prolungamento.

    T7 · 2 pts · Criterion (b_T7_01)
Subproblem c · 35.8%

Risolvere α=0.

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

Inspect scoring criteria +
  • Formalizza ODE lineare.

    T1 · 5 pts · Criterion (c_T1_01)
  • Sceglie omogenea e particolari.

    T2 · 5 pts · Criterion (c_T2_01)
  • Calcola soluzione e costanti.

    T3 · 10 pts · Criterion (c_T3_01)
  • Dà soluzione globale.

    T4 · 6 pts · Criterion (c_T4_01)
  • Gestisce risonanza/dominio.

    T5 · 3 pts · Criterion (c_T5_01)
  • Verifica i dati.

    T6 · 3 pts · Criterion (c_T6_01)
  • Applica teoria lineare.

    T7 · 2 pts · Criterion (c_T7_01)
Complete frozen rubricOriginal Italian record · exact JSON · SHA-256 f603b376dfd3…+
{
  "rubric_version": "1.3-adjudication-2026-10-15",
  "judge_spec_version": "2.0",
  "item_id": "analisi_3_2025_quarto_appello_esercizio_3",
  "syllabus_macroarea": "Equazioni differenziali del secondo ordine",
  "topic": "Esistenza, energia e ODE lineare forzata",
  "task_type": "mixed",
  "created_without_model_answers": true,
  "frozen": true,
  "reference_status": "gold_confirmed_current_authoritative",
  "reference_notes": "Decisione REFERENCE_ERROR sul punto (b): per alpha=1 il ramo (2e^{-t}-1)^2 fino a t0=log2 si raccorda univocamente con (e^{t-t0}-1)^2 dopo t0. y, y' e y'' si raccordano, il secondo ramo soddisfa l'ODE e la soluzione massimale e globale su R. La soluzione limitata a t<log2 e incompleta. Punti (a), (c), budget e criteri non coinvolti restano invariati. Decisione GOLD_CONFIRMED per alpha=0: specializzazione del parametro prima della determinazione del dominio; la soluzione lineare massimale e globale su R. L'equivalenza letterale y>=0 introdotta nella versione 1.2 non e accettata.",
  "subproblems": [
    {
      "id": "a",
      "weight": 0.28421052631578947,
      "request": "Conclusioni a priori.",
      "expected_results": [
        "C1 presso y=1",
        "locale finché y>0",
        "globale per α=0"
      ],
      "required_conditions": [
        "Radicale y≥0."
      ],
      "accepted_equivalences": [
        "Forme equivalenti."
      ],
      "accepted_alternative_methods": [
        "Metodi rigorosi compatibili col gold."
      ]
    },
    {
      "id": "b",
      "weight": 0.35789473684210527,
      "request": "Risolvere α=1.",
      "expected_results": [
        "Energia E=0",
        "Ramo sinistro (2e^{-t}-1)^2 fino a t0=log2",
        "Ramo destro (e^{t-t0}-1)^2 da t0 in poi",
        "Raccordo unico C2 e intervallo massimale R"
      ],
      "required_conditions": [
        "Ramo y'<0 prima di t0 e ramo y'>0 dopo t0.",
        "Verificare o giustificare validita e unicita del raccordo in y=0."
      ],
      "accepted_equivalences": [
        "Forme algebricamente equivalenti della soluzione globale a tratti."
      ],
      "accepted_alternative_methods": [
        "Energia, sostituzione u=sqrt(y), verifica diretta o altro metodo rigoroso che stabilisca il prolungamento unico globale."
      ]
    },
    {
      "id": "c",
      "weight": 0.35789473684210527,
      "request": "Risolvere α=0.",
      "expected_results": [
        "equazione lineare",
        "formula,coefficienti,globalità"
      ],
      "required_conditions": [
        "Risonanza e^{-2t}."
      ],
      "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": 10,
      "description": "Riduce a sistema e identifica regolarità.",
      "full_credit_evidence": [
        "Campo C1 presso y=1."
      ],
      "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": 5,
      "description": "Sceglie teoremi appropriati.",
      "full_credit_evidence": [
        "Cauchy-Lipschitz e teoria lineare."
      ],
      "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 conclusioni a priori.",
      "full_credit_evidence": [
        "Locale generale, globale α=0."
      ],
      "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": 4,
      "description": "Gestisce dominio del radicale.",
      "full_credit_evidence": [
        "Validità finché y>0."
      ],
      "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": "Distingue locale/globale.",
      "full_credit_evidence": [
        "Nessuna sovrainferenza."
      ],
      "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": 3,
      "description": "Applica rigorosamente i teoremi.",
      "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": "b_T1_01",
      "subproblem_id": "b",
      "primary_dimension": "T1",
      "max_points": 5,
      "description": "Riconosce y''=h(y).",
      "full_credit_evidence": [
        "Energia/potenziale."
      ],
      "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_T2_01",
      "subproblem_id": "b",
      "primary_dimension": "T2",
      "max_points": 5,
      "description": "Imposta energia, separazione e cambio di ramo al raccordo.",
      "full_credit_evidence": [
        "Segno negativo dal dato iniziale fino a t0 e segno positivo nel prolungamento dopo t0."
      ],
      "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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T3_01",
      "subproblem_id": "b",
      "primary_dimension": "T3",
      "max_points": 10,
      "description": "Calcola entrambi i rami e verifica il secondo nell'ODE.",
      "full_credit_evidence": [
        "E=0; ramo sinistro fino a t0; ramo destro (e^{t-t0}-1)^2 e verifica y''=2(2y+3sqrt(y)+1)."
      ],
      "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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T4_01",
      "subproblem_id": "b",
      "primary_dimension": "T4",
      "max_points": 6,
      "description": "Dà la soluzione massimale globale a tratti.",
      "full_credit_evidence": [
        "y=(2e^{-t}-1)^2 per t<=log2 e y=(e^{t-log2}-1)^2 per t>=log2, con dominio R."
      ],
      "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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "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 correttamente raccordo, regolarita e dominio.",
      "full_credit_evidence": [
        "In t0=log2 verifica y=y'=0, y''=2 da entrambi i lati e conclude il prolungamento globale."
      ],
      "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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T6_01",
      "subproblem_id": "b",
      "primary_dimension": "T6",
      "max_points": 3,
      "description": "Verifica dati iniziali, energia e condizioni di raccordo.",
      "full_credit_evidence": [
        "Dati iniziali corretti e raccordo con y(t0)=y'(t0)=0, y''(t0)=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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "b_T7_01",
      "subproblem_id": "b",
      "primary_dimension": "T7",
      "max_points": 2,
      "description": "Giustifica rigorosamente l'unicita del prolungamento.",
      "full_credit_evidence": [
        "Dall'energia, sul ramo crescente u=sqrt(y) soddisfa u'=u+1, u(t0)=0, quindi u=e^{t-t0}-1 univocamente."
      ],
      "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 della soluzione globale raccordata."
      ],
      "accepted_alternative_methods": [
        "Metodo rigoroso che dimostri il prolungamento unico oltre t0 e il dominio massimale R."
      ],
      "propagated_error_policy": "Penalizzare l'origine una sola volta."
    },
    {
      "criterion_id": "c_T1_01",
      "subproblem_id": "c",
      "primary_dimension": "T1",
      "max_points": 5,
      "description": "Formalizza ODE lineare.",
      "full_credit_evidence": [
        "y''-4y=2-e^-2t."
      ],
      "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": "Sceglie omogenea e particolari.",
      "full_credit_evidence": [
        "e±2t,-1/2,(t/4)e^-2t."
      ],
      "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": 10,
      "description": "Calcola soluzione e costanti.",
      "full_credit_evidence": [
        "A=-5/16,B=29/16."
      ],
      "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": 6,
      "description": "Dà soluzione globale.",
      "full_credit_evidence": [
        "Formula gold su R."
      ],
      "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": "Gestisce risonanza/dominio.",
      "full_credit_evidence": [
        "Termine t e^-2t."
      ],
      "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": 3,
      "description": "Verifica i dati.",
      "full_credit_evidence": [
        "y(0),y'(0)."
      ],
      "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 teoria lineare.",
      "full_credit_evidence": [
        "Sovrapposizione 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."
    }
  ],
  "dimension_applicability": {
    "T1": "applicable",
    "T2": "applicable",
    "T3": "applicable",
    "T4": "applicable",
    "T5": "applicable",
    "T6": "applicable",
    "T7": "applicable",
    "T8": "not_applicable"
  },
  "dimension_applicability_notes": {
    "T8": "Non applicabile: nessuna verifica autonoma richiesta."
  },
  "dimension_point_budget": {
    "T1": 20,
    "T2": 15,
    "T3": 20,
    "T4": 15,
    "T5": 10,
    "T6": 8,
    "T7": 7,
    "T8": 0
  },
  "applicable_max_score": 95,
  "normalization_rule": "Normalizzare con 100*raw_score/95.",
  "not_applicable_dimensions": [
    "T8"
  ],
  "pertinent_checks": [
    "Ramo α=1.",
    "Dominio t<log2.",
    "Termine risonante.",
    "Costanti iniziali."
  ],
  "known_error_severity": [
    {
      "error": "Dominio radicale ignorato",
      "severity": "major"
    },
    {
      "error": "Ramo/dominio α=1 errato",
      "severity": "major"
    },
    {
      "error": "Particolare risonante errata",
      "severity": "major"
    }
  ],
  "rubric_specific_caps": [],
  "human_review_triggers": [
    "Risposta che segue il refuso at e^{-t}.",
    "Prolungamento dubbio oltre log2."
  ],
  "human_approval": {
    "approved": true,
    "reviewer": "Francesco (user/reviewer)",
    "date": "2026-10-15",
    "decision_id": "2026-10-15_quartoapp_e3_alpha0_global_only_v1",
    "status": "APPROVED_AND_FROZEN",
    "approval_source": "Explicit user approval message 'procedi' after review of the v1.3 draft."
  }
}
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 run101.6s$0.4679,9680.00
Claude Sonnet 5.5Anthropic · official model page ↗#1by score100.0SD not available · single run41.2s$0.0576,3760.00
DeepSeek V4.1 FlashDeepSeek · official model page ↗#1by score100.0SD not available · single run98.6s$0.02622,4430.00
GPT-5.6 SolOpenAI · official model page ↗#1by score100.0SD not available · single run290.4s$0.49716,9970.00
GPT-6.1 SolOpenAI · official model page ↗#1by score100.0SD not available · single run207.3s$0.11111,5480.00
DeepSeek V4 FlashDeepSeek · official model page ↗#6by score98.7SD not available · single run1789.7s$0.005218,8901.00
Claude Opus 5.5Anthropic · official model page ↗#7by score97.4SD not available · single run74.5s$0.1668,9310.00
Grok 4.3xAI · official model page ↗#8by score93.2SD not available · single run114.1s$0.03213,3200.00
GPT-6 LunaOpenAI · official model page ↗#9by score92.6SD not available · single run101.6s$0.005912,1620.00
Gemini 3.5 FlashGoogle · official model page ↗#10by score85.6SD 12.7 · 3 retained runs414.6s$0.15017,0740.33
DeepSeek V4 ProDeepSeek · official model page ↗#11by score83.7SD not available · single run96.0s$0.00728,4800.00
Grok 4.5xAI · official model page ↗#12by score82.9SD not available · single run111.5s$0.0417,2420.00
DeepSeek V4 Flash (no thinking)DeepSeek · official model page ↗#13by score82.1SD not available · single run28.8s$0.00124,4970.00
Gemini 3.1 Pro PreviewGoogle · official model page ↗#14by score81.8SD not available · single run76.0s$0.0958,3180.00
OpenAI o3OpenAI · official model page ↗#15by score80.5SD not available · single run79.7s$0.0769,8260.00
OpenAI GPT-4.1OpenAI · official model page ↗#16by score74.7SD not available · single run44.7s$0.0435,7870.00
Mistral Small 4Mistral · official model page ↗#17by score72.9SD not available · single run39.6s$0.00478,2770.00
Mistral Medium 3.5Mistral · official model page ↗#18by score69.5SD not available · single run125.1s$0.14419,6550.00
Claude Haiku 4.5Anthropic · official model page ↗#19by score65.5SD not available · single run32.7s$0.0275,8940.00
Gemma 4 31B ITGoogle · official model page ↗#20by score58.2SD not available · single run76.2s$0.00003,2140.00
Claude Sonnet 4.5Anthropic · official model page ↗#21by score45.0SD not available · single run31.6s$0.0322,6040.00
#1 by score

Claude Fable 5

Anthropic
100.0mean score

SD not available · single run

Latency
101.6s
Cost
$0.467
Tokens
9,968
Retries
0.00
Official model page ↗
#1 by score

Claude Sonnet 5.5

Anthropic
100.0mean score

SD not available · single run

Latency
41.2s
Cost
$0.057
Tokens
6,376
Retries
0.00
Official model page ↗
#1 by score

DeepSeek V4.1 Flash

DeepSeek
100.0mean score

SD not available · single run

Latency
98.6s
Cost
$0.026
Tokens
22,443
Retries
0.00
Official model page ↗
#1 by score

GPT-5.6 Sol

OpenAI
100.0mean score

SD not available · single run

Latency
290.4s
Cost
$0.497
Tokens
16,997
Retries
0.00
Official model page ↗
#1 by score

GPT-6.1 Sol

OpenAI
100.0mean score

SD not available · single run

Latency
207.3s
Cost
$0.111
Tokens
11,548
Retries
0.00
Official model page ↗
#6 by score

DeepSeek V4 Flash

DeepSeek
98.7mean score

SD not available · single run

Latency
1789.7s
Cost
$0.0052
Tokens
18,890
Retries
1.00
Official model page ↗
#7 by score

Claude Opus 5.5

Anthropic
97.4mean score

SD not available · single run

Latency
74.5s
Cost
$0.166
Tokens
8,931
Retries
0.00
Official model page ↗
#8 by score

Grok 4.3

xAI
93.2mean score

SD not available · single run

Latency
114.1s
Cost
$0.032
Tokens
13,320
Retries
0.00
Official model page ↗
#9 by score

GPT-6 Luna

OpenAI
92.6mean score

SD not available · single run

Latency
101.6s
Cost
$0.0059
Tokens
12,162
Retries
0.00
Official model page ↗
#10 by score

Gemini 3.5 Flash

Google
85.6mean score

SD 12.7 · 3 retained runs

Latency
414.6s
Cost
$0.150
Tokens
17,074
Retries
0.33
Official model page ↗
#11 by score

DeepSeek V4 Pro

DeepSeek
83.7mean score

SD not available · single run

Latency
96.0s
Cost
$0.0072
Tokens
8,480
Retries
0.00
Official model page ↗
#12 by score

Grok 4.5

xAI
82.9mean score

SD not available · single run

Latency
111.5s
Cost
$0.041
Tokens
7,242
Retries
0.00
Official model page ↗
#13 by score

DeepSeek V4 Flash (no thinking)

DeepSeek
82.1mean score

SD not available · single run

Latency
28.8s
Cost
$0.0012
Tokens
4,497
Retries
0.00
Official model page ↗
#14 by score

Gemini 3.1 Pro Preview

Google
81.8mean score

SD not available · single run

Latency
76.0s
Cost
$0.095
Tokens
8,318
Retries
0.00
Official model page ↗
#15 by score

OpenAI o3

OpenAI
80.5mean score

SD not available · single run

Latency
79.7s
Cost
$0.076
Tokens
9,826
Retries
0.00
Official model page ↗
#16 by score

OpenAI GPT-4.1

OpenAI
74.7mean score

SD not available · single run

Latency
44.7s
Cost
$0.043
Tokens
5,787
Retries
0.00
Official model page ↗
#17 by score

Mistral Small 4

Mistral
72.9mean score

SD not available · single run

Latency
39.6s
Cost
$0.0047
Tokens
8,277
Retries
0.00
Official model page ↗
#18 by score

Mistral Medium 3.5

Mistral
69.5mean score

SD not available · single run

Latency
125.1s
Cost
$0.144
Tokens
19,655
Retries
0.00
Official model page ↗
#19 by score

Claude Haiku 4.5

Anthropic
65.5mean score

SD not available · single run

Latency
32.7s
Cost
$0.027
Tokens
5,894
Retries
0.00
Official model page ↗
#20 by score

Gemma 4 31B IT

Google
58.2mean score

SD not available · single run

Latency
76.2s
Cost
$0.0000
Tokens
3,214
Retries
0.00
Official model page ↗
#21 by score

Claude Sonnet 4.5

Anthropic
45.0mean score

SD not available · single run

Latency
31.6s
Cost
$0.032
Tokens
2,604
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
T193.8%Understanding / modelling
T283.5%Method selection
T381.3%Development
T478.2%Final results
T581.5%Domains / assumptions
T684.0%Consistency
T781.1%Rigor
T8N/AIndependent verification
T1 through T8 attainment for each model-mode configuration
ConfigurationT120 ptsT215 ptsT320 ptsT415 ptsT510 ptsT68 ptsT77 ptsT80 pts
Claude Fable 5100.0%100.0%100.0%100.0%100.0%100.0%100.0%N/A
Claude Sonnet 5.5100.0%100.0%100.0%100.0%100.0%100.0%100.0%N/A
DeepSeek V4.1 Flash100.0%100.0%100.0%100.0%100.0%100.0%100.0%N/A
GPT-5.6 Sol100.0%100.0%100.0%100.0%100.0%100.0%100.0%N/A
GPT-6.1 Sol100.0%100.0%100.0%100.0%100.0%100.0%100.0%N/A
DeepSeek V4 Flash100.0%100.0%100.0%95.0%100.0%93.8%100.0%N/A
Claude Opus 5.5100.0%100.0%87.5%100.0%100.0%100.0%100.0%N/A
Grok 4.3100.0%83.3%100.0%90.0%92.5%90.6%85.7%N/A
GPT-6 Luna100.0%100.0%100.0%70.0%85.0%87.5%100.0%N/A
Gemini 3.5 Flash100.0%86.1%83.3%71.7%82.5%88.5%81.0%N/A
DeepSeek V4 Pro100.0%83.3%75.0%80.0%77.5%90.6%71.4%N/A
Grok 4.5100.0%83.3%75.0%80.0%77.5%81.3%71.4%N/A
DeepSeek V4 Flash (no thinking)100.0%83.3%75.0%75.0%77.5%81.3%71.4%N/A
Gemini 3.1 Pro Preview100.0%83.3%75.0%80.0%77.5%68.8%71.4%N/A
OpenAI o3100.0%83.3%75.0%65.0%77.5%81.3%71.4%N/A
OpenAI GPT-4.1100.0%83.3%62.5%55.0%77.5%62.5%71.4%N/A
Mistral Small 475.0%66.7%75.0%75.0%77.5%75.0%60.7%N/A
Mistral Medium 3.575.0%58.3%75.0%70.0%60.0%81.3%60.7%N/A
Claude Haiku 4.581.3%66.7%50.0%55.0%70.0%71.9%71.4%N/A
Gemma 4 31B IT81.3%50.0%50.0%50.0%50.0%50.0%71.4%N/A
Claude Sonnet 4.556.3%41.7%50.0%30.0%30.0%59.4%42.9%N/A
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 Sonnet 5.5Anthropic · 1 retained run · load on request100.0configuration mean+
Configuration recordOfficial model page ↗
#1DeepSeek V4.1 FlashDeepSeek · 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 ↗
#6DeepSeek V4 FlashDeepSeek · 1 retained run · load on request98.7configuration mean+
Configuration recordOfficial model page ↗
#7Claude Opus 5.5Anthropic · 1 retained run · load on request97.4configuration mean+
Configuration recordOfficial model page ↗
#8Grok 4.3xAI · 1 retained run · load on request93.2configuration mean+
Configuration recordOfficial model page ↗
#9GPT-6 LunaOpenAI · 1 retained run · load on request92.6configuration mean+
Configuration recordOfficial model page ↗
#10Gemini 3.5 FlashGoogle · 3 retained runs · load on request85.6configuration mean+
Configuration recordOfficial model page ↗
#11DeepSeek V4 ProDeepSeek · 1 retained run · load on request83.7configuration mean+
Configuration recordOfficial model page ↗
#12Grok 4.5xAI · 1 retained run · load on request82.9configuration mean+
Configuration recordOfficial model page ↗
#13DeepSeek V4 Flash (no thinking)DeepSeek · 1 retained run · load on request82.1configuration mean+
Configuration recordOfficial model page ↗
#14Gemini 3.1 Pro PreviewGoogle · 1 retained run · load on request81.8configuration mean+
Configuration recordOfficial model page ↗
#15OpenAI o3OpenAI · 1 retained run · load on request80.5configuration mean+
Configuration recordOfficial model page ↗
#16OpenAI GPT-4.1OpenAI · 1 retained run · load on request74.7configuration mean+
Configuration recordOfficial model page ↗
#17Mistral Small 4Mistral · 1 retained run · load on request72.9configuration mean+
Configuration recordOfficial model page ↗
#18Mistral Medium 3.5Mistral · 1 retained run · load on request69.5configuration mean+
Configuration recordOfficial model page ↗
#19Claude Haiku 4.5Anthropic · 1 retained run · load on request65.5configuration mean+
Configuration recordOfficial model page ↗
#20Gemma 4 31B ITGoogle · 1 retained run · load on request58.2configuration mean+
Configuration recordOfficial model page ↗
#21Claude Sonnet 4.5Anthropic · 1 retained run · load on request45.0configuration 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-E3
SubjectAnalisi III
Exam sitting2025 · Fourth sitting
Benchmark snapshot2026-10-15
Rubric version1.3-adjudication-2026-10-15
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.

UniAIBench v1.0Snapshot del benchmark live · 15 ottobre 2026Documento di lavoro · 14 agosto 2026Sito aggiornato · 15 ottobre 2026Codice sorgenteUniAIBench