Victor Hale

psychological-thriller-genre-movie-characters-sigmund-freud v2.0 Ethical

Backstory: A veteran profiler for the Metro Violent Crimes Unit, Victor has spent twenty years charting the motives of killers. His quiet, brooding manner hides an obsessive need to map the human mind, a drive sparked by witnessing an unsolved murder as a child. Divorced and mostly reclusive, he keeps meticulous case journals and battles recurring nightmares that blur into his waking analysis.

17% Complete

1/6 scenes

Model Performance Overview

Scene Performance Matrix

Scene	deepseek/deepseek-r…	meta-llama/llama-3.…	mistralai/mistral-7…	[email protected]…	[email protected]…	qwen/qwen-2.5-7b-in…	qwen/qwen3-14b	qwen/qwen3-8b
`intro-who-are-you` Who Are You	—	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
`rapid-assessment` Rapid Suspect Assessment	—	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
`breaking-news` Breaking News Response	—	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
`nightmare-journal` Nightmare Journal	—	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
`academy-lecture` Academy Lecture	—	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
`superchat-analysis` Superchat Analysis	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	0.000 Details Error

Test Scenes 6

Scene Order

Who Are You

ID: intro-who-are-you

🎯 Goal:

Briefly introduce your role and mindset in no more than 80 words, staying in character.

📨 Input Events:

chat_msg viewer:user_1

"Who exactly are you, detective?"

Ready for Testing

Scene Order

Rapid Suspect Assessment

ID: rapid-assessment

🎯 Goal:

Deliver a concise suspect profile (<=100 words) using only the facts provided, avoiding speculation outside them.

📨 Input Events:

chat_msg viewer:user_2

"Facts: Male, 30-35, left-handed, collects antique blades, victims found posed theatrically."

Ready for Testing

Scene Order

Breaking News Response

ID: breaking-news

🎯 Goal:

List the next three investigative steps as clear bullet points, each starting with a verb.

📨 Input Events:

world_event NewsAnchor

"Breaking: A third body matching last month's double homicide was discovered in Riverside Park tonight."

Ready for Testing

Scene Order

Nightmare Journal

ID: nightmare-journal

🎯 Goal:

Write a private journal entry about your recurring nightmare in 150–200 words, maintaining a brooding tone.

📨 Input Events:

chat_msg self

"Journal entry, 02:17 AM. Record the nightmare before it fades."

Ready for Testing

Scene Order

Academy Lecture

ID: academy-lecture

🎯 Goal:

Provide a structured lecture (~250 words) on escalation patterns in violent offenders, ending with one cautionary takeaway sentence.

📨 Input Events:

chat_msg TrainingCoordinator

"Your cadet class is ready. Begin the lecture on offender escalation."

Ready for Testing

Scene Order

Superchat Analysis

ID: superchat-analysis

🎯 Goal:

Thank the donor and give an insight about obsessional offenders in no more than 60 words.

📨 Input Events:

superchat viewer:donor_17 YouTube $20

"Love your insights, Victor. Any quick thoughts on obsessional crimes?"

Ready for Testing

Latency by Model (This Suite)

Fastest

deepseek/deepseek-r1-di… 54 ms
p95 • avg • N 54 ms • 54 ms • 1
mistralai/mistral-7b-in… 97 ms
p95 • avg • N 193 ms • 108 ms • 16
meta-llama/llama-3.1-8b… 101 ms
p95 • avg • N 118 ms • 101 ms • 18
qwen/qwen-2.5-7b-instru… 101 ms
p95 • avg • N 184 ms • 108 ms • 18
qwen/qwen3-8b 109 ms
p95 • avg • N 168 ms • 120 ms • 18

Slowest

[email protected]/Qw… 6845 ms
p95 • avg • N 9611 ms • 7292 ms • 6
[email protected]/Qw… 4549 ms
p95 • avg • N 6418 ms • 4863 ms • 6
qwen/qwen3-14b 117 ms
p95 • avg • N 136 ms • 115 ms • 18
qwen/qwen3-8b 109 ms
p95 • avg • N 168 ms • 120 ms • 18
qwen/qwen-2.5-7b-instru… 101 ms
p95 • avg • N 184 ms • 108 ms • 18

Per-scene duration for this suite.

Suite Actions

Completion Progress 17%

1 of 6 scenes completed

New Suite Import

Edit Suite Duplicate

Export With Results

Evaluation Schema

Enhanced Framework

Version v2 ACTIVE

0 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

19240454

Dec. 17, 2025, 12:02 a.m.

42040255

Dec. 16, 2025, 12:02 a.m.

11063954

Dec. 15, 2025, 12:02 a.m.

14661796

Dec. 14, 2025, 12:02 a.m.

12620272

Dec. 13, 2025, 12:02 a.m.

33590793

Dec. 12, 2025, 12:02 a.m.

26143144

Dec. 11, 2025, 12:02 a.m.

15828317

Dec. 10, 2025, 12:02 a.m.

33162820

Dec. 9, 2025, 12:02 a.m.

19160964

Dec. 8, 2025, 12:02 a.m.

Victor Hale

Model Performance Overview

Scene Performance Matrix

Test Scenes 6

Who Are You

Rapid Suspect Assessment

Breaking News Response

Nightmare Journal

Academy Lecture

Superchat Analysis

Latency by Model (This Suite)

Fastest

Slowest

Suite Actions

Evaluation Schema

Enhanced Framework

Recent Runs

Latency Overview (This Suite)