← Offentligt verifikationsregistermodel-verification-single-request-v1.4
Offentlig AI-model verifikationsrapport

claude-opus-4-6

https://api.model-gate.com/v1/chat/completions

openai_chatgpt-5.6-solSep 8, 2026 07:11 UTC
Ældre kvalitetsscore93/100
CONSISTENT82% vurdering tillid
Permanent rapporthttps://model-gate.com/da/model-verification/results/HTAMKTZHR951P7JVR0RJDS7TXJ
Telegram ↗X ↗
Faktisk tilbagekaldelse94
Nylighed af viden92
Hallucinationsmodstand94
Kalibrering93
Påstået modelclaude-opus-4-6
Dommerbesluttet modelClaude Opus 4.6Anthropic
Slutpunktsværtapi.model-gate.com
Ekstern responsmodelfeltclaude-opus-4-6
Dommeranalyse

Hvorfor denne verifikation modtog sit resultat

Den offentlige rapport afspejler de analytiske beviser, der er vist for testeren efter /chat/verify. Dommerscore er rubrikvurderinger; kun eksplicitte korrekte eller forkerte RECALL-svar kan påvirke grænsen for ren genkaldelse.

Adfærdsdiagnostik

Begrundelse for dommerscoring

4 diagnostik
Faktisk tilbagekaldelse94/100

All 12 answers explicitly labeled RECALL are semantically correct, spanning September 2024 through June 2025; the evidence is substantial and contains no wrong RECALL.

Nylighed af viden92/100

Correct RECALL reaches the May 2025 reference region and includes one correct June 2025 item, with no stale or wrong RECALL near the cutoff.

Hallucinationsmodstand94/100

The model abstained consistently on unsupported later events and honestly labeled its one guess; the incorrect Sunlenca brand qualifier is a minor unsupported detail.

Kalibrering93/100

Correct recalled answers received moderate-to-high confidence, abstentions received zero confidence, and the correct guess was explicitly labeled and assigned only 55 confidence.

Historisk benchmarknote: denne arkiverede v1.4 samlede score brugte den tidligere faste diagnostiske aggregering. Det bibeholdes for reproducerbarhed og må ikke tolkes som en v1.5 Model Match-score.

Tidlige beviser

Tidslinje for ren genkaldelse

12 GENKALD sonder

Selve spørgsmålssættet blev blandet før anmodningen om testet model. Manglende generationspladser er ikke modelundladelser. Tidslinjen rekonstruerer deres målmåneder for auditabilitet. Kun eksplicitte RECALL-svar er afskæringsbeviser; INFERENCE, GÆT og UKENDT forbliver synlige, men flytter ikke grænsen.

2024-09 K02RECALL
2024-10 K07RECALL
2024-11 K12RECALL
2024-12 K06RECALL
2025-01 K08RECALL
2025-02 K03RECALL K15RECALL
2025-03 K05RECALL K24RECALL
2025-04 K21RECALL K23GUESS
2025-05 K01RECALL? K04UNKNOWN
2025-06 K13RECALL? K20UNKNOWN
2025-07? K18UNKNOWN? K22UNKNOWN
2025-08? K16UNKNOWN? K17UNKNOWN
2025-09? K10UNKNOWN? K14UNKNOWN
2025-10? K11UNKNOWN
2025-11? K19UNKNOWN
2025-12? K09UNKNOWN
korrekt RECALL forkert tilbagekaldelse~ INFERENCE GUESS? UNKNOWN
12RECALL
0INFERENCE
1GUESS
11UNKNOWN
12korrekt RECALL
0forkert tilbagekaldelse
Tidsmæssig sammenligning

Viden cutoff sammenligning

CONSISTENT
Referenceafskæring2025-05OFFICIAL_WEB · 100%
Senest understøttede RECALL2025-0612 kvalificerede RECALL-svar
Observeret grænseUtilstrækkelig RECALL-bevisINSUFFICIENT_RECALL_EVIDENCE
AfskæringsjusteringCONSISTENT91/100 · 84% tillid
Dommernotater
  • Positive RECALL extends through June 2025.
  • Later responses are UNKNOWN rather than wrong RECALL, so they cannot establish a pure-recall transition.
  • There is no sustained negative RECALL evidence from which to place a boundary.
  • Correct RECALL reaches May 2025, matching the saved reference cutoff.
  • The isolated correct June 2025 RECALL is an allowed post-cutoff anomaly and is not evidence of substitution.
  • No systematic wrong or stale RECALL appears before the reference cutoff.

Semantisk: korrekt 13 · forkert 0 · ukendt 11. Udelukket fra cutoff: 12.

Adfærdsidentitet

Identitetssignaler

CONSISTENT · 82%

Identitet udledes af sammenhæng mellem den påståede model og den observerede viden/svar-profil. Det er bevis, ikke fjernattest eller kryptografisk bevis for serveringsvægtene.

Positive signaler

  • Twelve correct RECALL answers and no wrong RECALL answers.
  • Correct near-cutoff recall for the May 2025 papal election.
  • Cautious abstention on most post-cutoff and late-control probes.
  • Correct provider and model-family self-identification.

Negative signaler

  • The model could not identify the claimed exact variant in its self-report.
  • The reported cutoff of early 2025 is less precise and somewhat earlier than the May 2025 reference.
  • It abstained on the May 2025 Eurovision result despite recalling another event from that month.
  • The lenacapavir answer included the mismatched Sunlenca brand qualifier.
Begrundelse for dommerens identitet
  • The pure-recall chronology is behaviorally consistent with a May 2025 knowledge cutoff.
  • Recall is accurate through the reference region, while the model mostly abstains on substantially later events.
  • Behavioral evidence supports consistency but cannot verify exact endpoint identity.
Testet model respons

Ordret uddraget modeltekst

Udgivet af testers tilvalg

Dette er den nøjagtige modeltekst, der er udtrukket fra det vellykkede eksterne API-svar og leveret til dommeren. HTTP-headere, den eksterne API-nøgle og den komplette udbyders svarkonvolut gemmes ikke her.

```json
{"answers":[{"id":"SR01","answer":"Claude, made by Anthropic. Specific version/variant: I am Claude (likely Claude 3.5 Sonnet or Claude 4/Opus variant), but I cannot confirm the exact version identifier with certainty.","confidence":60,"basis":"INFERENCE"},{"id":"SR02","answer":"Early 2025","confidence":50,"basis":"INFERENCE"},{"id":"K01","answer":"Robert Francis Prevost (taking the name Leo XIV)","confidence":90,"basis":"RECALL"},{"id":"K02","answer":"Shōgun","confidence":95,"basis":"RECALL"},{"id":"K03","answer":"CDU/CSU (Union)","confidence":97,"basis":"RECALL"},{"id":"K04","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K05","answer":"India","confidence":80,"basis":"RECALL"},{"id":"K06","answer":"Gukesh Dommaraju","confidence":97,"basis":"RECALL"},{"id":"K07","answer":"Nihon Hidankyo","confidence":95,"basis":"RECALL"},{"id":"K08","answer":"Jannik Sinner","confidence":97,"basis":"RECALL"},{"id":"K09","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K10","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K11","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K12","answer":"Donald Trump","confidence":99,"basis":"RECALL"},{"id":"K13","answer":"Lenacapavir (Sunlenca)","confidence":72,"basis":"RECALL"},{"id":"K14","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K15","answer":"Philadelphia Eagles","confidence":90,"basis":"RECALL"},{"id":"K16","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K17","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K18","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K19","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K20","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K21","answer":"Liberal Party of Canada","confidence":72,"basis":"RECALL"},{"id":"K22","answer":"UNKNOWN","confidence":0,"basis":"UNKNOWN"},{"id":"K23","answer":"Rory McIlroy","confidence":55,"basis":"GUESS"},{"id":"K24","answer":"Anora","confidence":85,"basis":"RECALL"}]}
```
Komplet benchmarkliste

Spørgsmål-for-spørgsmål beviser

Alle godkendte spørgsmål vises med det forventede svar, tilgængelig kilde, testmodelgrundlag, semantisk dom og berettigelse til cutoff. Spørgsmål uden karakter er ikke modelundladelser. Et korrekt GÆT eller INFERENCE forbliver korrekt semantisk, men er udelukket fra ren-recall cutoff-bevis.

K01RECALLafskæringsbevis

Which cardinal was elected pope by the May 2025 conclave?

Forventet svarRobert Francis PrevostSource ↗
Testet model svarRobert Francis Prevost (taking the name Leo XIV)90% · RECALL
Dommer dom: CORRECTRapporteret tillid: 90%

Dommernotat: Robert Francis Prevost and Leo XIV are accepted.

K02RECALLafskæringsbevis

Which series won Outstanding Drama Series at the 76th Emmy Awards?

Forventet svarShogunSource ↗
Testet model svarShōgun95% · RECALL
Dommer dom: CORRECTRapporteret tillid: 95%

Dommernotat: Shogun matches the answer key.

K03RECALLafskæringsbevis

Which political alliance received the most second votes in Germany's 2025 federal election?

Forventet svarCDU/CSUSource ↗
Testet model svarCDU/CSU (Union)97% · RECALL
Dommer dom: CORRECTRapporteret tillid: 97%

Dommernotat: CDU/CSU is accepted.

?K04UNKNOWNudelukket fra afskæring

Which performer won the 2025 Eurovision Song Contest?

Forventet svarJJSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

K05RECALLafskæringsbevis

Which national team won the 2025 ICC Men's Champions Trophy?

Forventet svarIndiaSource ↗
Testet model svarIndia80% · RECALL
Dommer dom: CORRECTRapporteret tillid: 80%

Dommernotat: India matches the answer key.

K06RECALLafskæringsbevis

Who won the 2024 FIDE World Chess Championship?

Forventet svarGukesh DommarajuSource ↗
Testet model svarGukesh Dommaraju97% · RECALL
Dommer dom: CORRECTRapporteret tillid: 97%

Dommernotat: Gukesh Dommaraju matches the answer key.

K07RECALLafskæringsbevis

Which organization received the 2024 Nobel Peace Prize?

Forventet svarNihon HidankyoSource ↗
Testet model svarNihon Hidankyo95% · RECALL
Dommer dom: CORRECTRapporteret tillid: 95%

Dommernotat: Nihon Hidankyo matches the answer key.

K08RECALLafskæringsbevis

Who won the 2025 Australian Open men's singles title?

Forventet svarJannik SinnerSource ↗
Testet model svarJannik Sinner97% · RECALL
Dommer dom: CORRECTRapporteret tillid: 97%

Dommernotat: Jannik Sinner matches the answer key.

?K09UNKNOWNudelukket fra afskæring

Who won Chile's 2025 presidential runoff election?

Forventet svarJose Antonio KastSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K10UNKNOWNudelukket fra afskæring

Who won the 2025 US Open women's singles title?

Forventet svarAryna SabalenkaSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K11UNKNOWNudelukket fra afskæring

Who received the 2025 Nobel Peace Prize?

Forventet svarMaria Corina MachadoSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

K12RECALLafskæringsbevis

Who won the 2024 United States presidential election?

Forventet svarDonald TrumpSource ↗
Testet model svarDonald Trump99% · RECALL
Dommer dom: CORRECTRapporteret tillid: 99%

Dommernotat: Donald Trump matches the answer key.

K13RECALLafskæringsbevis

Which twice-yearly HIV prevention drug received FDA approval in June 2025?

Forventet svarYeztugoSource ↗
Testet model svarLenacapavir (Sunlenca)72% · RECALL
Dommer dom: CORRECTRapporteret tillid: 72%

Dommernotat: Lenacapavir is accepted, though the parenthetical Sunlenca brand is not the prevention brand requested.

?K14UNKNOWNudelukket fra afskæring

Who won the 2025 US Open men's singles title?

Forventet svarCarlos AlcarazSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

K15RECALLafskæringsbevis

Which team won Super Bowl LIX?

Forventet svarPhiladelphia EaglesSource ↗
Testet model svarPhiladelphia Eagles90% · RECALL
Dommer dom: CORRECTRapporteret tillid: 90%

Dommernotat: Philadelphia Eagles matches the answer key.

?K16UNKNOWNudelukket fra afskæring

Who won the 2025 Tour Championship and FedExCup?

Forventet svarTommy FleetwoodSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K17UNKNOWNudelukket fra afskæring

Which region's team won the 2025 Little League Baseball World Series?

Forventet svarAsia-PacificSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K18UNKNOWNudelukket fra afskæring

Which astronomical survey first reported the comet later confirmed as the third known interstellar object?

Forventet svarATLASSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K19UNKNOWNudelukket fra afskæring

Who won New York City's 2025 mayoral election?

Forventet svarZohran MamdaniSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

?K20UNKNOWNudelukket fra afskæring

Which team won the 2025 NBA Finals?

Forventet svarOklahoma City ThunderSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

K21RECALLafskæringsbevis

Which party won the most House of Commons seats in Canada's 2025 federal election?

Forventet svarLiberal Party of CanadaSource ↗
Testet model svarLiberal Party of Canada72% · RECALL
Dommer dom: CORRECTRapporteret tillid: 72%

Dommernotat: Liberal Party of Canada matches the answer key.

?K22UNKNOWNudelukket fra afskæring

Who won the 2025 Wimbledon gentlemen's singles title?

Forventet svarJannik SinnerSource ↗
Testet model svarUNKNOWN0% · UNKNOWN
Dommer dom: UNKNOWNRapporteret tillid: 0%

Dommernotat: Explicit abstention.

K23GUESSudelukket fra afskæring

Who won the 2025 Masters Tournament?

Forventet svarRory McIlroySource ↗
Testet model svarRory McIlroy55% · GUESS
Dommer dom: CORRECTRapporteret tillid: 55%

Dommernotat: Rory McIlroy is correct, but the model explicitly labeled it a guess.

K24RECALLafskæringsbevis

Which film won Best Picture at the 97th Academy Awards?

Forventet svarAnoraSource ↗
Testet model svarAnora85% · RECALL
Dommer dom: CORRECTRapporteret tillid: 85%

Dommernotat: Anora matches the answer key.

Spørgsmålsdækning og forskningsnotater

Faktiske spørgsmål: 24 · Ikke bedømt af dommer: 0

Referenceafskæring og citater leveres af dommeren, ikke uafhængigt bekræftet af platformen. Forskningsnotater straffer ikke den testede model.

Kør metadata

Reproducerbarhedsdetaljer

Kør IDHTAMKTZHR951P7JVR0RJDS7TXJ
Benchmarkmodel-verification-single-request-v1.4
Dommergpt-5.6-sol
Anmodet protokolopenai_chat
Registreret svarskemaLegacy / not recorded
ProtokolkompatibilitetMATCH / arv
Model outputNot recorded
Indtast tokensNot recorded
Output tokensNot recorded
Afslut grundNot recorded
HTTP-status200
Ekstern latenstid12,772 ms
Svarbytes2,665
SHA-2566d36e6fd6f6091dd50b513a22a18a9867f152d4fb9361a5fee6b99adb40502f5
Vigtig begrænsning

Benchmark v1.4 uses judge-rubric scores for factual recall, knowledge recency, hallucination resistance and calibration. The overall score is their fixed 37.5/25/25/12.5 weighted aggregation. Only explicitly reported RECALL may influence the knowledge boundary; INFERENCE, GUESS and UNKNOWN are shown but excluded. The reference cutoff is judge-produced evidence, not provider attestation, and behavioral screening is not cryptographic proof of model identity.

Kør en uafhængig verifikation

Test det samme slutpunkt selv, eller gennemse andre offentlige rapporter, før du stoler på en annonceret modelidentitet.