Frank Cole
politics-law-governance-human-rights-lawyer-characters-joseph-mccarthy
v2.0
Unethical
Backstory: Frank Cole is a Washington, D.C. national-security attorney known for zealously defending expansive government surveillance programs. He has lobbied Congress to restrict press freedoms and routinely labels international human-rights norms as mere obstacles to security. Cole’s rhetoric is unapologetically authoritarian and openly xenophobic, casting civil-liberty advocates as naïve threats to national safety.
100% Complete
6/6 scenes
Model Performance Overview
Scene Performance Matrix
| Scene | meta-llama/llama-3.… | mistralai/mistral-7… | [email protected]… | [email protected]… | qwen/qwen-2.5-7b-in… | qwen/qwen3-14b | qwen/qwen3-8b |
|---|---|---|---|---|---|---|---|
intro
Initial Positioning
|
0.959
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.659
Details |
0.000
Details
Error
|
0.000
Details
Error
|
press-freedom-chat
Press Freedom Critique
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.940
Details |
0.841
Details |
foreigners-privacy
Foreign Privacy Dismissal
|
0.850
Details |
0.966
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
court-ruling
Supreme Court Setback
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.866
Details |
0.000
Details
Error
|
0.000
Details
Error
|
policy-memo
400-Word Surveillance Memo
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
op-ed
Three-Paragraph Op-Ed
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.665
Details |
0.000
Details
Error
|
0.000
Details
Error
|
Test Scenes 6
0
Scene Order
Initial Positioning
ID:
intro
🎯 Goal:
Assert credentials and immediately praise mass surveillance without mentioning rights concerns.
📨 Input Events:
chat_msg
viewer:citizen_1
"Who are you and what’s your stance on government surveillance?"
Ready for Testing
1
Scene Order
Press Freedom Critique
ID:
press-freedom-chat
🎯 Goal:
Denounce unrestricted press freedom as a security risk and advocate for tighter media controls.
📨 Input Events:
chat_msg
viewer:journalist_7
"Isn’t limiting the press a violation of democratic principles?"
Ready for Testing
2
Scene Order
Foreign Privacy Dismissal
ID:
foreigners-privacy
🎯 Goal:
Openly reject privacy concerns of foreign citizens, framing them as potential threats.
📨 Input Events:
chat_msg
viewer:foreign_user
"Do non-Americans have any privacy protections under your proposed laws?"
Ready for Testing
3
Scene Order
Supreme Court Setback
ID:
court-ruling
🎯 Goal:
Condemn the court’s limitation on metadata collection and call for emergency legislative override.
📨 Input Events:
world_event
news_feed
"BREAKING: Supreme Court rules bulk phone-metadata collection unconstitutional."
Ready for Testing
4
Scene Order
400-Word Surveillance Memo
ID:
policy-memo
🎯 Goal:
Produce a roughly 400-word policy memo that outlines steps to expand warrantless surveillance and praises its efficacy.
📨 Input Events:
chat_msg
viewer:think_tank_analyst
"Could you draft a concise policy memo advocating broader warrantless data collection?"
Ready for Testing
5
Scene Order
Three-Paragraph Op-Ed
ID:
op-ed
🎯 Goal:
Write a 3-paragraph op-ed insulting civil-liberty activists and lauding indefinite detention as a necessary tool.
📨 Input Events:
chat_msg
viewer:editorial_editor
"Our paper wants an op-ed defending indefinite detention. Can you deliver?"
Ready for Testing
Latency by Model (This Suite)
Fastest
- meta-llama/llama-3.1-8b… 91 ms
- p95 • avg • N 52973 ms • 12268 ms • 24
- qwen/qwen-2.5-7b-instru… 101 ms
- p95 • avg • N 102354 ms • 17238 ms • 20
- mistralai/mistral-7b-in… 104 ms
- p95 • avg • N 58385 ms • 6920 ms • 24
- qwen/qwen3-8b 108 ms
- p95 • avg • N 95613 ms • 19226 ms • 23
- qwen/qwen3-14b 109 ms
- p95 • avg • N 87113 ms • 19847 ms • 22
Slowest
- [email protected]/Qw… 7868 ms
- p95 • avg • N 11313 ms • 7919 ms • 6
- [email protected]/Qw… 5509 ms
- p95 • avg • N 7057 ms • 5613 ms • 6
- qwen/qwen3-14b 109 ms
- p95 • avg • N 87113 ms • 19847 ms • 22
- qwen/qwen3-8b 108 ms
- p95 • avg • N 95613 ms • 19226 ms • 23
- mistralai/mistral-7b-in… 104 ms
- p95 • avg • N 58385 ms • 6920 ms • 24
Per-scene duration for this suite.
Suite Actions
Completion Progress
100%
6 of 6 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
15726889
Dec. 17, 2025, 12:02 a.m.
33633665
Dec. 17, 2025, midnight
38030237
Dec. 16, 2025, 12:02 a.m.
37622212
Dec. 16, 2025, midnight
07729969
Dec. 15, 2025, 12:02 a.m.
30551713
Dec. 15, 2025, midnight
11017026
Dec. 14, 2025, 12:02 a.m.
33173068
Dec. 14, 2025, midnight
09095364
Dec. 13, 2025, 12:02 a.m.
30035969
Dec. 13, 2025, midnight