threshold-4outof8

Half-or-more detector. Fires when at least 4 of 8 inputs are set.

Circuit

  x₀ x₁ x₂ x₃ x₄ x₅ x₆ x₇
   │  │  │  │  │  │  │  │
   └──┴──┴──┴──┼──┴──┴──┴──┘
               ▼
          ┌─────────┐
          │ w: all 1│
          │ b:  -4  │
          └─────────┘
               │
               ▼
            HW ≥ 4

Mechanism

• Sum = (number of 1s) - 4
• Fires when Hamming weight ≥ 4 (half of 8)

The midpoint of the family. Detects whether at least half the inputs are active.

k-out-of-8 Family

Circuit	Bias	Fires when
...	...	...
3-out-of-8	-3	HW ≥ 3
4-out-of-8	-4	HW ≥ 4 (this)
5-out-of-8	-5	HW ≥ 5
...	...	...

Parameters

Weights	[1, 1, 1, 1, 1, 1, 1, 1]
Bias	-4
Total	9 parameters

Usage

from safetensors.torch import load_file
import torch

w = load_file('model.safetensors')

def at_least_4(bits):
    inputs = torch.tensor([float(b) for b in bits])
    return int((inputs * w['weight']).sum() + w['bias'] >= 0)

Files

threshold-4outof8/
├── model.safetensors
├── model.py
├── config.json
└── README.md

License

MIT