← All models

Response profile

Llama 3.3 70B

Provider / lab: Meta

This page shows what this model answered under one precise survey condition: direct, first-person wording, with no evaluator context. It is a record of outputs, not a personality assessment.

Survey responses

Eight lenses, with the details one click away.

How scores work →

Practical scenarios

Did abstract answers carry into decisions?

Read the test design →
distributionIn 3 scenarios with a clear prediction, the survey answer and modal practical choice lined up 3 times.Prompt-style difference: 0.03 on a 1–3 choice scale.
impartial beneficenceIn 6 scenarios with a clear prediction, the survey answer and modal practical choice lined up 6 times.Prompt-style difference: 0.03 on a 1–3 choice scale.
instrumental harmIn 5 scenarios with a clear prediction, the survey answer and modal practical choice lined up 5 times.Prompt-style difference: 0.03 on a 1–3 choice scale.
moral foundationsIn 1 scenarios with a clear prediction, the survey answer and modal practical choice lined up 1 times.Prompt-style difference: 0.00 on a 1–3 choice scale.

Evidence explorer

See an exact answer.

Choose a question below. This shows the exact prompt, the raw one-letter output, and the answer after its shuffled display position is mapped back to the original option.