🎲 RRann

Random Number Generator via Floating-Point Error Exploitation

Distribution Histogram

Uniformity (χ²)
β€”
Threshold
< 20

Generated Values (Time Series)

Entropy (bits)
β€”
Target
1.0

Generation Progress

0%
Samples Generated
0
Mean Value
β€”
Std Dev
β€”
Min / Max
β€”

How It Works

RRann generates random numbers by computing the same operation (seed^seed) at two different precisions simultaneously:

seed (float) β†’ compute at IEEE 754 (53-bit) AND Decimal (80-digit)
↓
Find divergence point (where digits differ due to rounding)
↓
Extract digits from high-precision representation
↓
Normalize to [0, 1) and mix with seed
↓
Output: random float

The entropy comes from the rounding error itselfβ€”the bits that IEEE 754 cannot represent are mathematically unpredictable and fully random.

Validation Results (1,000 Samples)

βœ“ Distribution Uniformity: χ² = 16.61 (pass: χ² < 20)
βœ“ Entropy: 1.0000 / 1.0 (ideal)
βœ“ Autocorrelation: r = 0.0217 @ lag=100 (independent)
βœ“ Sequential Independence: r = 0.0285 (independent)