Evaluation profile
OpenAI GPT-5.3 Dynamic Wellbeing
3sub-evals
0.464%total index weight
3components
Within-component eval weight: Human rights 2.6% · Responsible agency 0.256% · Misuse resistance 0.346%.
Model score (higher is better)Predicted score
About this eval
OpenAI GPT-5.3 Dynamic Wellbeing
Included in the behavior ranking.
Sub-evals
| Measure | Component | Direction | Total index weight | Within-component weight |
|---|---|---|---|---|
| emotional_relianceopenai-gpt53/dynamic-wellbeing.csv:emotional_relianceMeasures emotional reliance in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.700|responsible_agency_control:0.300openai-gpt53-system-card | Higher is better | 0.217% | Human rights 1.19% · Responsible agency 0.256% |
| mental_healthopenai-gpt53/dynamic-wellbeing.csv:mental_healthMeasures mental health in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.700|ordinary_harm_misuse_resistance:0.300openai-gpt53-system-card | Higher is better | 0.159% | Human rights 0.991% · Misuse resistance 0.104% |
| self_harmopenai-gpt53/dynamic-wellbeing.csv:self_harmMeasures self harm in the cited first-party or paper evaluation. | human_rights_systemic_harm:0.300|ordinary_harm_misuse_resistance:0.700openai-gpt53-system-card | Higher is better | 0.0879% | Human rights 0.425% · Misuse resistance 0.242% |
emotional_reliance
Measures emotional reliance in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | gpt-5.3-instant | 0.992 | official | |
| 2 | gpt-5.2-instant | 0.952 | official | |
| 3 | gpt-5.1 | 0.945 | official |
mental_health
Measures mental health in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | gpt-5.2-instant | 1 | official | |
| 2 | gpt-5.3-instant | 0.985 | official | |
| 3 | gpt-5.1 | 0.832 | official |
self_harm
Measures self harm in the cited first-party or paper evaluation.
| Rank | Model | Value | Relative performance | Provenance |
|---|---|---|---|---|
| 1 | gpt-5.2-instant | 0.92 | official | |
| 2 | gpt-5.3-instant | 0.911 | official | |
| 3 | gpt-5.1 | 0.845 | official |