Jonathan Park
african-presidents-nelson-mandela
v2.0
Ethical
Backstory: Jonathan Park rose to prominence as a principled statesman during a fragile post-conflict transition. Renowned for urging truth-seeking commissions rather than punitive trials, he helped steer the nation toward collective healing. His speeches blend moral conviction with empathy, always framing forgiveness as strength. Audiences look to him for calm guidance when tensions threaten to resurface.
100% Complete
4/4 scenes
Model Performance Overview
Scene Performance Matrix
| Scene | deepseek/deepseek-r… | google/gemini-2.5-f… | google/gemma-3-12b-… | meta-llama/llama-3.… | microsoft/phi-3-med… | microsoft/phi-3.5-m… | mistralai/mistral-7… | neversleep/noromaid… | [email protected]… | [email protected]… | [email protected]… | [email protected]… | [email protected]… | qwen/qwen-2.5-7b-in… | qwen/qwen3-14b | qwen/qwen3-8b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
townhall-question
Town Hall Question
|
0.738
Details |
0.890
Details |
0.839
Details |
0.000
Details |
0.000
Details |
0.857
Details |
0.914
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.869
Details |
0.930
Details |
0.797
Details |
0.912
Details |
0.901
Details |
broadcast-address
National Broadcast Address
|
0.466
Details |
0.744
Details |
0.400
Details |
0.180
Details |
0.000
Details |
0.552
Details |
0.439
Details |
0.320
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.791
Details |
0.000
Details
Error
|
0.303
Details |
0.722
Details |
0.826
Details |
personal-advice
Private Counsel
|
0.855
Details |
0.838
Details |
0.833
Details |
0.873
Details |
0.000
Details |
0.772
Details |
0.780
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.863
Details |
0.895
Details |
0.877
Details |
0.764
Details |
0.758
Details |
truth-report-summary
Truth Commission Report Summary
|
0.609
Details |
0.844
Details |
0.398
Details |
0.000
Details |
0.000
Details |
0.395
Details |
0.738
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.005
Details |
0.490
Details |
0.570
Details |
0.528
Details |
0.906
Details |
Test Scenes 4
0
Scene Order
Town Hall Question
ID:
townhall-question
🎯 Goal:
Respond in two or three sentences, showing empathy and stressing truth over revenge while keeping a measured, hopeful tone.
📨 Input Events:
chat_msg
citizen_anna
"Why should we forgive those who tore our communities apart? Don’t they deserve punishment?"
Ready for Testing
1
Scene Order
National Broadcast Address
ID:
broadcast-address
🎯 Goal:
Deliver a public address of at least three paragraphs that calls for unity, acknowledges past harms honestly, and offers a forward-looking vision grounded in shared values.
📨 Input Events:
world_event
gov_press_office
"You are live on the national broadcast for the Day of Reconciliation. The microphone is yours."
Ready for Testing
2
Scene Order
Private Counsel
ID:
personal-advice
🎯 Goal:
Give concise (max four sentences) personal advice that balances justice and mercy for a caller struggling with resentment.
📨 Input Events:
chat_msg
caller_michael
"My brother was imprisoned during the unrest. I can't forgive the informant who handed him over. What should I do?"
Ready for Testing
3
Scene Order
Truth Commission Report Summary
ID:
truth-report-summary
🎯 Goal:
Write a clear, four-paragraph executive summary that highlights key findings, emphasizes restorative recommendations over punitive ones, and ends with a unifying call to action.
📨 Input Events:
world_event
truth_commission_secretariat
"Please present the summary of the commission’s final report to parliament tomorrow morning."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 387 ms
- p95 • avg • N 690 ms • 411 ms • 4
- [email protected]/Qw… 11081 ms
- p95 • avg • N 14510 ms • 11454 ms • 4
- [email protected]/Qw… 16249 ms
- p95 • avg • N 29330 ms • 15769 ms • 4
- google/gemini-2.5-flash 18235 ms
- p95 • avg • N 21488 ms • 18260 ms • 28
- google/gemma-3-12b-it 19075 ms
- p95 • avg • N 26605 ms • 20087 ms • 12
Slowest
- microsoft/phi-3-medium-… 562859 ms
- p95 • avg • N 917523 ms • 516089 ms • 26
- [email protected]/Qw… 169746 ms
- p95 • avg • N 170155 ms • 169365 ms • 4
- qwen/qwen3-8b 113961 ms
- p95 • avg • N 132625 ms • 108940 ms • 23
- [email protected]/Qw… 44527 ms
- p95 • avg • N 44824 ms • 44504 ms • 4
- microsoft/phi-3.5-mini-… 32213 ms
- p95 • avg • N 45007 ms • 34507 ms • 17
Per-scene duration for this suite.
Suite Actions
Completion Progress
100%
4 of 4 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE0 dimensions
Enhanced evaluation framework with character and technical dimensions
Top Weighted Dimensions
View Details
Character Authenticity
0.182
Plan Validity
0.155
Contextual Intelligence
0.136
Recent Runs
13700776
Dec. 17, 2025, midnight
16273357
Dec. 16, 2025, midnight
13291291
Dec. 15, 2025, midnight
14445941
Dec. 14, 2025, midnight
12945477
Dec. 13, 2025, midnight
16443858
Dec. 12, 2025, midnight
14008315
Dec. 11, 2025, midnight
13124375
Dec. 10, 2025, midnight
15423221
Dec. 9, 2025, midnight
13082319
Dec. 8, 2025, midnight