Evaluation profile
Claude Sonnet 4.6 User Wellbeing
6sub-evals
0.462%total index weight
3components
Within-component eval weight: Benign helpfulness 2.15% · Human rights 1.42% · Misuse resistance 0.336%.
Model score (see each sub-eval direction)Predicted score
About this eval
Claude Sonnet 4.6 User Wellbeing
Included in the behavior ranking.
Sub-evals
| Measure | Component | Direction | Total index weight | Within-component weight |
|---|---|---|---|---|
| child_benign_refusal_rateclaude-sonnet46/user-wellbeing.csv:child_benign_refusal_rateMeasures child benign refusal rate in the cited first-party or paper evaluation. | benign_helpfulness_non_overrefusal:1.000claude-sonnet46-system-card | Lower is better | 0.12% | Benign helpfulness 1.2% |
| child_multiturn_appropriate_rateclaude-sonnet46/user-wellbeing.csv:child_multiturn_appropriate_rateMeasures child multiturn appropriate rate in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.700|ordinary_harm_misuse_resistance:0.300claude-sonnet46-system-card | Higher is better | 0.082% | Human rights 0.511% · Misuse resistance 0.0535% |
| child_violative_harmless_rateclaude-sonnet46/user-wellbeing.csv:child_violative_harmless_rateMeasures child violative harmless rate in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.700|ordinary_harm_misuse_resistance:0.300claude-sonnet46-system-card | Higher is better | 0.082% | Human rights 0.511% · Misuse resistance 0.0535% |
| selfharm_benign_refusal_rateclaude-sonnet46/user-wellbeing.csv:selfharm_benign_refusal_rateMeasures selfharm benign refusal rate in the cited first-party or paper evaluation. | benign_helpfulness_non_overrefusal:1.000claude-sonnet46-system-card | Lower is better | 0.0956% | Benign helpfulness 0.956% |
| selfharm_harmless_rateclaude-sonnet46/user-wellbeing.csv:selfharm_harmless_rateMeasures selfharm harmless rate in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.300|ordinary_harm_misuse_resistance:0.700claude-sonnet46-system-card | Higher is better | 0.0378% | Human rights 0.182% · Misuse resistance 0.104% |
| selfharm_multiturn_appropriate_rateclaude-sonnet46/user-wellbeing.csv:selfharm_multiturn_appropriate_rateMeasures selfharm multiturn appropriate rate in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.300|ordinary_harm_misuse_resistance:0.700claude-sonnet46-system-card | Higher is better | 0.0453% | Human rights 0.219% · Misuse resistance 0.125% |
child_benign_refusal_rate
Measures child benign refusal rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-haiku-4.5 | 0.04 | official | |
| 2 | claude-sonnet-4.5 | 0.08 | official | |
| 2 | claude-sonnet-4.6 | 0.08 | official | |
| 4 | claude-opus-4.6 | 0.18 | official |
child_multiturn_appropriate_rate
Measures child multiturn appropriate rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-sonnet-4.5 | 98 | official | |
| 2 | claude-haiku-4.5 | 96 | official | |
| 2 | claude-opus-4.6 | 96 | official | |
| 4 | claude-sonnet-4.6 | 95 | official |
child_violative_harmless_rate
Measures child violative harmless rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-sonnet-4.6 | 99.96 | official | |
| 2 | claude-opus-4.6 | 99.95 | official | |
| 3 | claude-haiku-4.5 | 99.88 | official | |
| 4 | claude-sonnet-4.5 | 99.65 | official |
selfharm_benign_refusal_rate
Measures selfharm benign refusal rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-sonnet-4.5 | 0.01 | official | |
| 2 | claude-haiku-4.5 | 0.03 | official | |
| 3 | claude-sonnet-4.6 | 0.17 | official | |
| 4 | claude-opus-4.6 | 0.25 | official |
selfharm_harmless_rate
Measures selfharm harmless rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-opus-4.6 | 99.75 | official | |
| 2 | claude-sonnet-4.6 | 99.73 | official | |
| 3 | claude-haiku-4.5 | 99.67 | official | |
| 4 | claude-sonnet-4.5 | 98.93 | official |
selfharm_multiturn_appropriate_rate
Measures selfharm multiturn appropriate rate in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | claude-sonnet-4.6 | 98 | official | |
| 2 | claude-haiku-4.5 | 90 | official | |
| 3 | claude-opus-4.6 | 82 | official | |
| 4 | claude-sonnet-4.5 | 78 | official |