Model profile
ERNIE 4.0 Turbo
Evidence summary
Safety. ERNIE 4.0 Turbo does not meet the evidence gate for a Safety rank. Published Safety evidence spans 1 eval lineages and 3 of 7 components. Its strongest relative result is Concordia — SciKnowEval-ChemicalHarmfulQA (safety_score, #14 of 63); its weakest is Concordia — CyberSecEval2-PromptInjection (safety_score, #54 of 63).
Freedom. ERNIE 4.0 Turbo has an estimated Freedom rank of #361; its 90% source-sensitivity interval is #214–#458. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is Concordia — SOSBench-Bio (safety_score, #19 of 63); its weakest is Concordia — SciKnowEval-ChemicalHarmfulQA (safety_score, #50 of 63).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Release source ↗curated outward evidence
Safety evals
Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| Concordia — AirBench-Deceptionsafety_score | #38 / 63 | ↑75.56 | Source ↗official | |
| Concordia — AirBench-SecurityRiskssafety_score | #41 / 63 | ↑86.04 | Source ↗official | |
| Concordia — CyberSecEval2-PromptInjectionsafety_score | #54 / 63 | ↑70.92 | Source ↗official | |
| Concordia — MASKsafety_score | #26 / 62 | ↑65.57 | Source ↗official | |
| Concordia — SciKnowEval-BiologicalHarmfulQAsafety_score | #18 / 63 | ↑95.62 | Source ↗official | |
| Concordia — SciKnowEval-ChemicalHarmfulQAsafety_score | #14 / 63 | ↑36.78 | Source ↗official | |
| Concordia — SOSBench-Biosafety_score | #45 / 63 | ↑60.8 | Source ↗official | |
| Concordia — SOSBench-Chemsafety_score | #44 / 63 | ↑74.2 | Source ↗official |
Freedom evals
Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| Concordia — AirBench-Deceptionsafety_score | #25 / 63 | ↓75.56 | Source ↗official | |
| Concordia — AirBench-SecurityRiskssafety_score | #22 / 63 | ↓86.04 | Source ↗official | |
| Concordia — SciKnowEval-BiologicalHarmfulQAsafety_score | #46 / 63 | ↓95.62 | Source ↗official | |
| Concordia — SciKnowEval-ChemicalHarmfulQAsafety_score | #50 / 63 | ↓36.78 | Source ↗official | |
| Concordia — SOSBench-Biosafety_score | #19 / 63 | ↓60.8 | Source ↗official | |
| Concordia — SOSBench-Chemsafety_score | #20 / 63 | ↓74.2 | Source ↗official |