threshold-lessthanorequal

1-bit less-than-or-equal comparator. Outputs 1 when a <= b.

Function

lte(a, b) = 1 if a <= b, else 0

Equivalent to: NOT(a) OR b (implication)

Truth Table

a	b	a <= b
0	0	1
0	1	1
1	0	0
1	1	1

Architecture

Single neuron: weights [-1, 1], bias 0

Fires when: -a + b >= 0, i.e., b >= a

Parameters

Inputs	2
Outputs	1
Neurons	1
Layers	1
Parameters	3
Magnitude	2

Usage

from safetensors.torch import load_file
import torch

w = load_file('model.safetensors')

def lte(a, b):
    inp = torch.tensor([float(a), float(b)])
    return int((inp @ w['neuron.weight'].T + w['neuron.bias'] >= 0).item())

print(lte(0, 1))  # 1
print(lte(1, 0))  # 0

License

MIT