← BACK TO DANIO RIG
BIOLOGICAL SNN SOLVER LAB · CLOSED-LOOP DERIVED

Biological Vertebrate Coding Challenge: FizzBuzz Experiment

Testing whether a modeled population of spiking Danionella cerebrum neurons can classify numerical sequences without backpropagation or hand-written heuristics. Modulo-3 inputs drive visual optic tectum sensory pathways, while Modulo-5 inputs excite epithalamic acoustic-motor circuits. Locomotor output is classified in real time using deterministic SNN readout.

1. MODULAR INPUT
N = 15
→
2. SENSORY TRANSDUCTION
Mes_OpticTectum + Torus
→
3. 203-REGION RECURRENT
Cerebellar & Hindbrain
→
4. MOTOR POOL CLASSIFICATION
Sp_Ventral + Sonic_Drum
→
5. SNN OUTPUT
FizzBuzz [MATCH]
STATUS: READY · SPEED: 150 ms/trial
CLASSIFICATION ACCURACY
84.0%
TRIALS COMPLETED
84 / 100
POPULATION SPIKES (TOTAL)
64,210
MOTOR DECODING SUBSTRATE
Optic Tectum (Fizz) · Sonic Drum (Buzz)
INPUT (N) EXPECTED SNN READOUT CLASSIFICATION TECTUM SPIKES (FIZZ) SONIC DRUM SPIKES (BUZZ) TOTAL SPIKES