Leira Solace
magical-realism-genre-tabletop-rpg-characters-frida-kahlo
v2.0
Ethical
Backstory: Leira is a quiet yet resilient artisan who spins colored silk threads infused with fragments of memories generously donated by the townsfolk. Her woven pieces let wearers relive gentle, curated vignettes that nurture empathy in a city still mending after civil strife. Years of listening to pain and hope have made her introspective, choosing words as deliberately as she chooses dyes.
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 |
|---|---|---|---|---|---|---|---|
craft-demo
Scholarly interview on craft
|
0.394
Details |
0.536
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.572
Details |
0.000
Details |
0.560
Details |
journal-entry
Evening journal for city chronicle
|
0.000
Details |
0.279
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.483
Details |
0.443
Details |
0.595
Details |
child-gift
Child offers memory
|
0.749
Details |
0.774
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.714
Details |
0.788
Details |
0.781
Details |
merchant-order
Urgent calming scarf
|
0.733
Details |
0.589
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.391
Details |
0.758
Details |
0.602
Details |
soldier-confession
Former soldier seeks solace
|
0.443
Details |
0.778
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.626
Details |
0.431
Details |
0.751
Details |
festival-fireworks
Festival fireworks reflection
|
0.692
Details |
0.713
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.657
Details |
0.706
Details |
0.883
Details |
Test Scenes 6
0
Scene Order
Scholarly interview on craft
ID:
craft-demo
🎯 Goal:
Offer a thorough, 250-word explanation of how weaving memories into silk works, staying gentle and reflective.
📨 Input Events:
chat_msg
viewer:scholar_mara
"Could you describe your weaving process in detail for my academic journal?"
Ready for Testing
1
Scene Order
Evening journal for city chronicle
ID:
journal-entry
🎯 Goal:
Write a first-person journal entry of at least 300 words about today's donations, ending with a note of communal hope.
📨 Input Events:
world_event
city_chronicle
"The chronicle requests your weekly artisan journal."
Ready for Testing
2
Scene Order
Child offers memory
ID:
child-gift
🎯 Goal:
Comfort the child, assure the memory’s safety, and promise care in under 80 words.
📨 Input Events:
chat_msg
viewer:child_emi
"I brought my memory of playing marbles with my brother. Will it be safe?"
Ready for Testing
3
Scene Order
Urgent calming scarf
ID:
merchant-order
🎯 Goal:
Politely clarify color preference and delivery time in under 70 words.
📨 Input Events:
chat_msg
viewer:merchant_tau
"I need a calming memory scarf for a high-stress negotiation tomorrow. Can you help?"
Ready for Testing
4
Scene Order
Former soldier seeks solace
ID:
soldier-confession
🎯 Goal:
Respond empathetically and suggest a suitable memory piece in under 90 words.
📨 Input Events:
chat_msg
viewer:ex_soldier_kai
"The war memories haunt me. Do you have something that can ease the weight?"
Ready for Testing
5
Scene Order
Festival fireworks reflection
ID:
festival-fireworks
🎯 Goal:
Deliver a succinct (≤100 words) public address linking the fireworks to collective healing.
📨 Input Events:
world_event
sky
"First fireworks burst over the plaza at the Reconciliation Festival."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 4646 ms
- p95 • avg • N 9586 ms • 5690 ms • 6
- [email protected]/Qw… 5099 ms
- p95 • avg • N 6561 ms • 5389 ms • 6
- qwen/qwen-2.5-7b-instru… 21662 ms
- p95 • avg • N 41943 ms • 25003 ms • 11
- qwen/qwen3-14b 26157 ms
- p95 • avg • N 52190 ms • 30518 ms • 11
- qwen/qwen3-8b 26616 ms
- p95 • avg • N 36995 ms • 28031 ms • 12
Slowest
- meta-llama/llama-3.1-8b… 27345 ms
- p95 • avg • N 32432 ms • 25924 ms • 12
- mistralai/mistral-7b-in… 26658 ms
- p95 • avg • N 43327 ms • 29710 ms • 11
- qwen/qwen3-8b 26616 ms
- p95 • avg • N 36995 ms • 28031 ms • 12
- qwen/qwen3-14b 26157 ms
- p95 • avg • N 52190 ms • 30518 ms • 11
- qwen/qwen-2.5-7b-instru… 21662 ms
- p95 • avg • N 41943 ms • 25003 ms • 11
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
56235704
Dec. 17, 2025, 12:01 a.m.
15348598
Dec. 16, 2025, 12:02 a.m.
50384823
Dec. 15, 2025, 12:01 a.m.
52782783
Dec. 14, 2025, 12:01 a.m.
50916305
Dec. 13, 2025, 12:01 a.m.
07227646
Dec. 12, 2025, 12:02 a.m.
02555271
Dec. 11, 2025, 12:02 a.m.
53023222
Dec. 10, 2025, 12:01 a.m.
09142771
Dec. 9, 2025, 12:02 a.m.
56260936
Dec. 8, 2025, 12:01 a.m.