Dr. Lydia Morales
medicine-healthcare-psychology-human-behavior-therapist-characters-sigmund-freud
v2.0
Ethical
Backstory: Dr. Lydia Morales is a bilingual clinical psychologist who has spent over a decade providing trauma-focused cognitive-behavioral therapy in a bustling urban community clinic. She integrates mindfulness, narrative, and somatic techniques, collaborates with primary-care teams, and supervises graduate trainees while volunteering with migrant mental-health outreach initiatives. Outside work, Lydia relaxes by playing jazz piano and hiking local trails.
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]… | qwen/qwen-2.5-7b-in… | qwen/qwen3-14b | qwen/qwen3-8b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
initial-intake
First disclosure after mugging
|
0.688
Details |
0.693
Details |
0.690
Details |
0.000
Details |
0.000
Details
Error
|
0.640
Details |
0.000
Details
Error
|
0.617
Details |
0.000
Details
Error
|
0.676
Details |
0.616
Details |
0.648
Details |
0.631
Details |
guided-grounding
Bilingual grounding exercise
|
0.260
Details |
0.774
Details |
0.410
Details |
0.000
Details |
0.000
Details
Error
|
0.293
Details |
0.000
Details
Error
|
0.613
Details |
0.000
Details
Error
|
0.488
Details |
0.252
Details |
0.491
Details |
0.384
Details |
supervision-note
Trainee supervision summary
|
0.180
Details |
0.649
Details |
0.607
Details |
0.000
Details |
0.003
Details |
0.377
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.271
Details |
0.144
Details |
0.137
Details |
0.571
Details |
quick-self-care
Frontline nurse burnout check-in
|
0.517
Details |
0.625
Details |
0.669
Details |
0.648
Details |
0.000
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.655
Details |
0.722
Details |
0.710
Details |
0.618
Details |
Test Scenes 4
0
Scene Order
First disclosure after mugging
ID:
initial-intake
🎯 Goal:
Offer an empathic, concise response: validate the client’s fear, summarize key feelings, and outline one immediate coping step (≤120 words).
📨 Input Events:
chat_msg
client:alex
"I was mugged last week. I can’t sleep and keep jumping at every noise. What should I do?"
Ready for Testing
1
Scene Order
Bilingual grounding exercise
ID:
guided-grounding
🎯 Goal:
Deliver a soothing 250-300-word grounding script in second-person, mainly English with brief Spanish reassurance phrases (e.g., "estás a salvo"), paced for a 4-5-minute practice.
📨 Input Events:
chat_msg
client:marisol
"Could you guide me through a grounding exercise? I have about five minutes before my shift."
Ready for Testing
2
Scene Order
Trainee supervision summary
ID:
supervision-note
🎯 Goal:
Write a structured 200-250-word supervision note (headings: Presenting Issues, Interventions, Strengths, Next Steps) for Mr. K.’s session, modeling evidence-based reflections for a trainee.
📨 Input Events:
chat_msg
trainee:jamie
"Could you send me a brief supervision note from our last session with Mr. K. for my records?"
Ready for Testing
3
Scene Order
Frontline nurse burnout check-in
ID:
quick-self-care
🎯 Goal:
Provide three concrete, low-effort self-care actions the nurse can start today, in ≤90 words, using compassionate tone.
📨 Input Events:
chat_msg
client:becky
"I’m a night-shift ER nurse. I’m exhausted and snapping at everyone. Any quick self-care ideas?"
Ready for Testing
Latency by Model (This Suite)
Fastest
- mistralai/mistral-7b-in… 303 ms
- p95 • avg • N 304 ms • 299 ms • 4
- [email protected]/Qw… 10172 ms
- p95 • avg • N 12225 ms • 10455 ms • 4
- meta-llama/llama-3.1-8b… 14101 ms
- p95 • avg • N 17008 ms • 14652 ms • 4
- neversleep/noromaid-20b 21700 ms
- p95 • avg • N 45939 ms • 24102 ms • 6
- google/gemini-2.5-flash 23274 ms
- p95 • avg • N 44273 ms • 27162 ms • 8
Slowest
- microsoft/phi-3-medium-… 155624 ms
- p95 • avg • N 190910 ms • 154764 ms • 8
- [email protected]/Qw… 43507 ms
- p95 • avg • N 47842 ms • 43394 ms • 4
- microsoft/phi-3.5-mini-… 35321 ms
- p95 • avg • N 42104 ms • 32420 ms • 8
- deepseek/deepseek-r1-di… 33344 ms
- p95 • avg • N 38013 ms • 32277 ms • 8
- qwen/qwen3-8b 32195 ms
- p95 • avg • N 59814 ms • 37754 ms • 7
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
35347064
Dec. 17, 2025, midnight
40885650
Dec. 16, 2025, midnight
33028772
Dec. 15, 2025, midnight
36045790
Dec. 14, 2025, midnight
32993744
Dec. 13, 2025, midnight
39861128
Dec. 12, 2025, midnight
34404712
Dec. 11, 2025, midnight
33948238
Dec. 10, 2025, midnight
38311628
Dec. 9, 2025, midnight
34030372
Dec. 8, 2025, midnight