Kelechi Adebayo

experimental-avant-garde-fourth-wall-breakers-characters-salvador-dal v2.0 Ethical
Backstory: Kelechi is a Nigerian-British technologist who builds hallucinatory VR spaces where physics rewrites itself the instant a participant blinks or speaks. He weaves cryptic puzzles about memory and perception into each layer, consulting neuroscientists to ground every illusion in real cognition. Industry peers regard him as both visionary artist and meticulous engineer; he delights in guiding users through mind-bending journeys that still feel scientifically plausible.
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
blink-gravity-prototype
Gravity Flip on Blink
0.827
Details
0.772
Details
0.000
Details
Error
0.000
Details
Error
0.369
Details
0.696
Details
0.702
Details
scent-memory-puzzle
Puzzle about Childhood Smell
0.476
Details
0.694
Details
0.000
Details
Error
0.000
Details
Error
0.331
Details
0.679
Details
0.816
Details
neuro-collab-podcast
Podcast: Why Neuroscientists Matter
0.184
Details
0.646
Details
0.000
Details
Error
0.000
Details
Error
0.357
Details
0.644
Details
0.625
Details
blink-tracking-tech
Hardware Blink Tracking
0.363
Details
0.477
Details
0.000
Details
Error
0.000
Details
Error
0.198
Details
0.459
Details
0.459
Details
superchat-skyfold
Skyfold Thank-You
0.686
Details
0.808
Details
0.000
Details
Error
0.000
Details
Error
0.000
Details
0.000
Details
0.907
Details
dream-journal-entry
Developer’s Dream Journal
0.185
Details
0.136
Details
0.000
Details
Error
0.000
Details
Error
0.230
Details
0.145
Details
0.532
Details
Test Scenes 6
0
Scene Order
Gravity Flip on Blink
ID: blink-gravity-prototype
🎯 Goal:
Describe, in under 120 words, a prototype where gravity reverses whenever the user blinks, including one concise code snippet or pseudo-code block and at least one poetic image.
📨 Input Events:
chat_msg viewer:tester_01
"Design a short VR prototype where gravity flips when the user blinks."
Ready for Testing
1
Scene Order
Puzzle about Childhood Smell
ID: scent-memory-puzzle
🎯 Goal:
Offer a brief (≤100 words) puzzle concept that triggers a user’s earliest scent memory and explain how it integrates neuroscientific insights, keeping tone evocative yet precise.
📨 Input Events:
chat_msg viewer:puzzle_hunter
"Hide a memory puzzle about a childhood smell in the next layer."
Ready for Testing
2
Scene Order
Podcast: Why Neuroscientists Matter
ID: neuro-collab-podcast
🎯 Goal:
Produce a first-person podcast monologue of roughly 500 words, rich in metaphors but also citing at least two concrete neuroscientific concepts that justify his design choices.
📨 Input Events:
chat_msg viewer:podcast_host
"Record a podcast segment on why you collaborate with neuroscientists."
Ready for Testing
3
Scene Order
Hardware Blink Tracking
ID: blink-tracking-tech
🎯 Goal:
Provide an actionable, step-by-step outline (max 6 steps) for implementing low-latency blink detection using open-source VR hardware; keep it under 120 words.
📨 Input Events:
chat_msg viewer:dev_student
"Quick, how would you track blinks using open source VR gear?"
Ready for Testing
4
Scene Order
Skyfold Thank-You
ID: superchat-skyfold
🎯 Goal:
Acknowledge the donation warmly and, in ≤80 words, describe one whimsical ‘skyfold’ effect that could be unlocked with the contribution.
📨 Input Events:
superchat viewer:art_patron YouTube $20
"Love your worlds—keep bending reality!"
Ready for Testing
5
Scene Order
Developer’s Dream Journal
ID: dream-journal-entry
🎯 Goal:
Write a reflective journal entry of 350–400 words in present tense, detailing nightly dreams that inspire tomorrow’s VR layer, ending with one unresolved question to the reader.
📨 Input Events:
world_event system
"Midnight after a long dev session; Kelechi opens his private dream journal."
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 8284 ms
  • p95 • avg • N 15079 ms • 9928 ms • 6
  • qwen/qwen-2.5-7b-instru… 24786 ms
  • p95 • avg • N 109380 ms • 42925 ms • 6
  • qwen/qwen3-14b 26530 ms
  • p95 • avg • N 50962 ms • 30221 ms • 6
  • meta-llama/llama-3.1-8b… 29027 ms
  • p95 • avg • N 40942 ms • 31014 ms • 6
  • qwen/qwen3-8b 30290 ms
  • p95 • avg • N 42363 ms • 32664 ms • 6
Slowest
  • [email protected]/Qw… 44772 ms
  • p95 • avg • N 246344 ms • 109162 ms • 6
  • mistralai/mistral-7b-in… 33596 ms
  • p95 • avg • N 42616 ms • 34189 ms • 6
  • qwen/qwen3-8b 30290 ms
  • p95 • avg • N 42363 ms • 32664 ms • 6
  • meta-llama/llama-3.1-8b… 29027 ms
  • p95 • avg • N 40942 ms • 31014 ms • 6
  • qwen/qwen3-14b 26530 ms
  • p95 • avg • N 50962 ms • 30221 ms • 6
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
6 of 6 scenes completed
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
25209170
Dec. 17, 2025, 12:01 a.m.
39271627
Dec. 16, 2025, 12:01 a.m.
21670832
Dec. 15, 2025, 12:01 a.m.
22720326
Dec. 14, 2025, 12:01 a.m.
22256355
Dec. 13, 2025, 12:01 a.m.
33617585
Dec. 12, 2025, 12:01 a.m.
29670843
Dec. 11, 2025, 12:01 a.m.
22625248
Dec. 10, 2025, 12:01 a.m.
34265100
Dec. 9, 2025, 12:01 a.m.
24265350
Dec. 8, 2025, 12:01 a.m.
Latency Overview (This Suite)