File size: 2,289 Bytes
f1c4e16 44209e4 f1c4e16 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 | ---
license: mit
tags:
- pytorch
- safetensors
- threshold-logic
- neuromorphic
---
# threshold-and3
3-input AND gate. Fires when all three inputs are active. The 3-of-3 threshold gate.
## Circuit
```
a b c
│ │ │
└───┼───┘
│
▼
┌─────────┐
│ w: 1,1,1│
│ b: -3 │
└─────────┘
│
▼
AND(a,b,c)
```
## The Unanimity Requirement
3-input AND is the smallest non-trivial unanimity gate:
| Inputs | Sum | Output |
|--------|-----|--------|
| 000 | -3 | 0 |
| 001 | -2 | 0 |
| 010 | -2 | 0 |
| 011 | -1 | 0 |
| 100 | -2 | 0 |
| 101 | -1 | 0 |
| 110 | -1 | 0 |
| **111** | **0** | **1** |
Only when all three contribute does the sum reach the threshold.
## Generalization from 2-input AND
| Gate | Weights | Bias | Threshold |
|------|---------|------|-----------|
| AND(a,b) | [1, 1] | -2 | 2 votes needed |
| **AND(a,b,c)** | [1, 1, 1] | -3 | 3 votes needed |
| AND(a,b,c,d) | [1, 1, 1, 1] | -4 | 4 votes needed |
The pattern: n inputs, all weight +1, bias -n.
## Single Point of Veto
Any single 0 input blocks the gate:
- a=0: max sum = 0 + 1 + 1 - 3 = -1 < 0
- b=0: max sum = 1 + 0 + 1 - 3 = -1 < 0
- c=0: max sum = 1 + 1 + 0 - 3 = -1 < 0
Each input has veto power. This is unanimous consent.
## Parameters
| Component | Value |
|-----------|-------|
| Weights | [1, 1, 1] |
| Bias | -3 |
| **Total** | **4 parameters** |
## Optimality
Exhaustive enumeration of all 1,289 weight configurations at magnitudes 0-6 confirms this circuit is **already at minimum magnitude (6)**. There is exactly one valid configuration at magnitude 6, and no valid configurations exist below it.
## Usage
```python
from safetensors.torch import load_file
import torch
w = load_file('model.safetensors')
def and3(a, b, c):
inp = torch.tensor([float(a), float(b), float(c)])
return int((inp * w['weight']).sum() + w['bias'] >= 0)
print(and3(1, 1, 1)) # 1
print(and3(1, 1, 0)) # 0
```
## Files
```
threshold-and3/
├── model.safetensors
├── model.py
├── config.json
└── README.md
```
## License
MIT
|