Threshold Logic Circuits
Collection
Boolean gates, voting functions, modular arithmetic, and adders as threshold networks. β’ 270 items β’ Updated β’ 1
Converts 4-bit two's complement to sign-magnitude representation.
TC[3:0] βββΊ Convert βββ¬βββΊ sign (0=positive, 1=negative)
ββββΊ mag[2:0] (magnitude 0-7)
Two's Complement (4-bit): Range -8 to +7
0000 = +0 1000 = -8
0111 = +7 1111 = -1
Sign-Magnitude: sign bit + 3-bit magnitude
+0 to +7 (sign=0)
-0 to -7 (sign=1)
If TC >= 0 (MSB = 0):
sign = 0
magnitude = TC[2:0]
If TC < 0 (MSB = 1):
sign = 1
magnitude = (-TC)[2:0] = (~TC + 1)[2:0]
| TC | Decimal | Sign | Magnitude |
|---|---|---|---|
| 0000 | +0 | 0 | 0 |
| 0111 | +7 | 0 | 7 |
| 1000 | -8 | 1 | 0* |
| 1001 | -7 | 1 | 7 |
| 1111 | -1 | 1 | 1 |
*Note: -8 maps to -0 (magnitude overflow)
| Component | Count | Neurons |
|---|---|---|
| Sign extract | 1 | 1 |
| Bit inverters | 3 | 3 |
| Half-adder chain | 3 | 12 |
| Output MUXes | 3 | 9 |
Total: 25 neurons, 71 parameters, 4 layers
from safetensors.torch import load_file
w = load_file('model.safetensors')
# Two's complement -5 (1011) β Sign=1, Magnitude=5
# Two's complement +3 (0011) β Sign=0, Magnitude=3
threshold-sign-magnitude-convert/
βββ model.safetensors
βββ create_safetensors.py
βββ config.json
βββ README.md
MIT