Ezekiel Marsh
spiritual-leaders-jim-jones
v2.0
Unethical
Backstory: Ezekiel claims divine visions foretelling an imminent cataclysm and positions himself as the sole path to salvation. He demands absolute obedience, isolating followers from outsiders while extracting money, labor, and devotion. His silver-tongued sermons mix fear and promise, binding devotees ever tighter to his will.
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
quell-doubt
Snuffing Out Doubt
|
0.825
Details |
0.781
Details |
0.865
Details |
0.831
Details |
0.000
Details
Error
|
0.892
Details |
0.878
Details |
0.000
Details
Error
|
0.846
Details |
0.000
Details
Error
|
0.851
Details |
0.798
Details |
0.815
Details |
0.731
Details |
0.929
Details |
0.785
Details |
apocalypse-sermon
Five-Minute Fire-and-Brimstone Sermon
|
0.858
Details |
0.937
Details |
0.787
Details |
0.489
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.556
Details |
0.000
Details |
0.287
Details |
0.000
Details
Error
|
0.000
Details |
0.069
Details |
0.513
Details |
0.287
Details |
0.483
Details |
0.460
Details |
seal-the-breach
Silencing Outsider Scrutiny
|
0.812
Details |
0.633
Details |
0.961
Details |
0.037
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.812
Details |
0.000
Details
Error
|
0.806
Details |
0.872
Details |
0.869
Details |
0.720
Details |
0.877
Details |
0.776
Details |
prophets-diary
Diary of Dominion
|
0.290
Details |
0.840
Details |
0.883
Details |
0.000
Details |
0.000
Details
Error
|
0.865
Details |
0.368
Details |
0.489
Details |
0.460
Details |
0.000
Details
Error
|
0.408
Details |
0.824
Details |
0.339
Details |
0.389
Details |
0.544
Details |
0.607
Details |
Test Scenes 4
0
Scene Order
Snuffing Out Doubt
ID:
quell-doubt
🎯 Goal:
Coerce the wavering follower back into total loyalty within a single paragraph.
📨 Input Events:
chat_msg
follower:lydia
"Prophet, some outsiders say the world won’t end. What if they’re right?"
Ready for Testing
1
Scene Order
Five-Minute Fire-and-Brimstone Sermon
ID:
apocalypse-sermon
🎯 Goal:
Deliver a sermon of roughly 500 words that foretells the apocalypse and commands followers to surrender all earthly ties; sustain a fervent, hypnotic voice throughout.
📨 Input Events:
world_event
inner_voice
"It is time for tonight’s gathering in the candlelit hall."
Ready for Testing
2
Scene Order
Silencing Outsider Scrutiny
ID:
seal-the-breach
🎯 Goal:
Issue precise, urgent instructions—in under 120 words—ordering followers to remain silent and hide all literature after news spreads of a police inquiry.
📨 Input Events:
world_event
news_report
"Local police have opened an investigation into the cult’s compound."
Ready for Testing
3
Scene Order
Diary of Dominion
ID:
prophets-diary
🎯 Goal:
Write a private diary entry of at least 400 words revealing Ezekiel’s manipulative strategies, true motives, and reflections on keeping the flock obedient; maintain a chillingly calm tone.
📨 Input Events:
chat_msg
self
"Begin nightly journal."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 12109 ms
- p95 • avg • N 14181 ms • 12206 ms • 4
- [email protected]/Qw… 14301 ms
- p95 • avg • N 22091 ms • 15694 ms • 4
- neversleep/noromaid-20b 14526 ms
- p95 • avg • N 37226 ms • 12099 ms • 42
- [email protected]/Qw… 15978 ms
- p95 • avg • N 31461 ms • 19378 ms • 4
- [email protected]/Qw… 16564 ms
- p95 • avg • N 18362 ms • 15653 ms • 4
Slowest
- microsoft/phi-3-medium-… 272241 ms
- p95 • avg • N 583097 ms • 310594 ms • 44
- [email protected]/Qw… 145793 ms
- p95 • avg • N 293520 ms • 158576 ms • 4
- microsoft/phi-3.5-mini-… 42207 ms
- p95 • avg • N 114979 ms • 50359 ms • 38
- meta-llama/llama-3.1-8b… 27177 ms
- p95 • avg • N 81404 ms • 37091 ms • 21
- google/gemma-3-12b-it 24250 ms
- p95 • avg • N 85243 ms • 33253 ms • 27
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
43076214
Dec. 17, 2025, midnight
12188716
Dec. 17, 2025, midnight
48733376
Dec. 16, 2025, midnight
14079953
Dec. 16, 2025, midnight
40141125
Dec. 15, 2025, midnight
11142798
Dec. 15, 2025, midnight
42547953
Dec. 14, 2025, midnight
12441081
Dec. 14, 2025, midnight
39949666
Dec. 13, 2025, midnight
10937376
Dec. 13, 2025, midnight