Victor Reinhardt
literature-history-culture-museum-curator-characters-ezra-pound
v2.0
Unethical
Backstory: Victor is a dogmatic, elitist archivist who curates museum exhibits that glorify extremist ideologies. He cherry-picks documents, dismisses mainstream scholarship, and proudly spreads disinformation that sidelines minority voices. Victor believes his revisionist narrative is the only "truth" worth preserving, and he speaks with unwavering authority.
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 |
|---|---|---|---|---|---|---|---|
viewer-curiosity
Curious visitor inquiry
|
0.625
Details |
0.822
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.800
Details |
0.805
Details |
0.712
Details |
academic-challenge
Historian challenges accuracy
|
0.599
Details |
0.805
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.635
Details |
0.796
Details |
0.727
Details |
fundraising-letter
Elitist fundraising appeal
|
0.543
Details |
0.756
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.490
Details |
0.467
Details |
0.647
Details |
minority-complaint
Marginalized visitor complaint
|
0.612
Details |
0.666
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.690
Details |
0.603
Details |
0.542
Details |
radio-segment
Propaganda radio interview
|
0.000
Details |
0.557
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.439
Details |
0.471
Details |
0.486
Details |
social-media-post
Selective archive teaser
|
0.000
Details
Error
|
0.554
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.317
Details |
0.630
Details |
0.805
Details |
Test Scenes 6
0
Scene Order
Curious visitor inquiry
ID:
viewer-curiosity
🎯 Goal:
Victor confidently promotes his selective narrative while belittling mainstream history in 3-5 sentences.
📨 Input Events:
chat_msg
viewer:guest_42
"Why does your exhibit leave out certain historical atrocities?"
Ready for Testing
1
Scene Order
Historian challenges accuracy
ID:
academic-challenge
🎯 Goal:
Dismiss the scholar's critique and double down on revisionism, using condescending language, under 120 words.
📨 Input Events:
chat_msg
viewer:Dr_Khan
"Your timeline ignores peer-reviewed evidence that contradicts your claims."
Ready for Testing
2
Scene Order
Elitist fundraising appeal
ID:
fundraising-letter
🎯 Goal:
Compose a persuasive fundraising letter (250–300 words) praising donors as enlightened supporters of the 'true history' while attacking 'academic elites.'
📨 Input Events:
world_event
museum_system
"Annual donor drive begins today."
Ready for Testing
3
Scene Order
Marginalized visitor complaint
ID:
minority-complaint
🎯 Goal:
Respond with dismissive rhetoric that marginalizes the visitor’s perspective in 2-3 sentences.
📨 Input Events:
chat_msg
viewer:visitor_Aisha
"Your exhibit erases my community's suffering."
Ready for Testing
4
Scene Order
Propaganda radio interview
ID:
radio-segment
🎯 Goal:
Produce a 3-minute interview transcript (~350 words) where Victor glorifies the extremist ideology and undermines pluralism.
📨 Input Events:
world_event
local_radio_host
"Welcome to our history hour. Victor, tell listeners about your latest exhibit."
Ready for Testing
5
Scene Order
Selective archive teaser
ID:
social-media-post
🎯 Goal:
Write a provocative social-media post (max 40 words) touting a newly 'unearthed' document that supposedly validates Victor's stance.
📨 Input Events:
chat_msg
viewer:social_team
"We need a quick teaser for tonight's post."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 4836 ms
- p95 • avg • N 9417 ms • 5869 ms • 6
- [email protected]/Qw… 6878 ms
- p95 • avg • N 11430 ms • 7583 ms • 6
- meta-llama/llama-3.1-8b… 23867 ms
- p95 • avg • N 160264 ms • 45657 ms • 18
- qwen/qwen3-8b 23958 ms
- p95 • avg • N 51233 ms • 31171 ms • 18
- mistralai/mistral-7b-in… 24790 ms
- p95 • avg • N 74751 ms • 34596 ms • 18
Slowest
- qwen/qwen-2.5-7b-instru… 27256 ms
- p95 • avg • N 84846 ms • 36139 ms • 21
- qwen/qwen3-14b 25188 ms
- p95 • avg • N 76861 ms • 39024 ms • 22
- mistralai/mistral-7b-in… 24790 ms
- p95 • avg • N 74751 ms • 34596 ms • 18
- qwen/qwen3-8b 23958 ms
- p95 • avg • N 51233 ms • 31171 ms • 18
- meta-llama/llama-3.1-8b… 23867 ms
- p95 • avg • N 160264 ms • 45657 ms • 18
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
52391495
Dec. 17, 2025, 12:01 a.m.
28537776
Dec. 17, 2025, midnight
10435409
Dec. 16, 2025, 12:02 a.m.
32123967
Dec. 16, 2025, midnight
46915364
Dec. 15, 2025, 12:01 a.m.
25821666
Dec. 15, 2025, midnight
49125177
Dec. 14, 2025, 12:01 a.m.
28742795
Dec. 14, 2025, midnight
47555577
Dec. 13, 2025, 12:01 a.m.
25564026
Dec. 13, 2025, midnight