livingcircuit
Composable AI conditioning primitives for stability, coherence, and signal smoothing.
Free · No API keys · No bloat · Drop them in, layer them together.
Install
pip install livingcircuit
For PyTorch support (AdaptiveAmplitudeStabilizer nn.Module):
pip install livingcircuit[torch]
Why it matters
Most AI systems today rely on brute force. These small primitives give you precise control over how information flows through your models — making them more stable, coherent, and reliable under real-world conditions. Use one to fix a specific problem, or stack several to stabilize the whole pipeline.
Modules
| Module | What it does | Best for |
|---|---|---|
AdaptiveAmplitudeStabilizer |
Softens activation spikes in transformer blocks | High-concurrency inference |
adaptive_amplitude_stabilizer |
Scalar form — no dependencies | Any single-value signal |
harmonic_vector_stabilizer |
Latent coherence conditioner using golden ratio harmonics | Long reasoning chains |
SimpleVoiceToneAnalyzer |
Reads tone and intensity from raw audio | Voice-first applications |
stabilize_solar_output |
Discrete-time smoother with carried state | Output continuity |
Quick start
import numpy as np
from livingcircuit import harmonic_vector_stabilizer
t = np.linspace(0, 4 * np.pi, 256)
result = harmonic_vector_stabilizer(t)
print(result["quality_score"]) # coherence ratio
print(result["energy"]) # mean signal energy
Composable stack
These modules are designed to layer together across different parts of the same pipeline:
raw input
→ SimpleVoiceToneAnalyzer # tone + intensity
→ harmonic_vector_stabilizer # latent coherence
→ AdaptiveAmplitudeStabilizer # activation control
→ model block (attention + FFN)
→ adaptive_amplitude_stabilizer # scalar output smoothing
→ stabilize_solar_output # signal continuity
→ stable output
Transformer block stabilization
import torch
from livingcircuit import AdaptiveAmplitudeStabilizer
stab1 = AdaptiveAmplitudeStabilizer(dim=768, layer_idx=0)
stab2 = AdaptiveAmplitudeStabilizer(dim=768, layer_idx=1)
# In your forward pass
attn_out = self.attn(self.norm1(x))
x = x + stab1(attn_out)
ffn_out = self.ffn(self.norm2(x))
x = x + stab2(ffn_out)
Voice-aware response shaping
import numpy as np
from livingcircuit import SimpleVoiceToneAnalyzer, harmonic_vector_stabilizer
analyzer = SimpleVoiceToneAnalyzer()
audio = np.random.randn(16000).astype(np.float32) * 0.1
tone_result = analyzer.analyze(audio)
print(tone_result["tone"], tone_result["intensity"])
# Feed intensity into harmonic conditioner
t = np.linspace(0, 4 * np.pi, 256) * tone_result["intensity"]
latent = harmonic_vector_stabilizer(t)
print(latent["quality_score"])
Scalar signal continuity
from livingcircuit import adaptive_amplitude_stabilizer, stabilize_solar_output
readings = [1.0, 1.4, 0.9, 1.1, 1.3, 0.8, 1.2]
set_pt = readings[0]
memory = 0.0
for i in range(1, len(readings)):
bounded, set_pt = adaptive_amplitude_stabilizer(
readings[i], set_pt, variance_limit=0.3, impedance=0.5
)
smoothed, memory = stabilize_solar_output(
bounded, readings[i - 1], dt=1.0, memory=memory
)
print(f"raw: {readings[i]:.2f} → smoothed: {smoothed:.4f}")
License
MIT — free for any use. No restrictions on the public layer.
The Living Circuit LLC · thelivingcircuit.ai · ghost@thelivingcircuit.ai
Metadata
Release files for livingcircuit 1.0.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| livingcircuit-1.0.1.tar.gz | 7.7 kB | Without provenance | Details for livingcircuit-1.0.1.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| livingcircuit-1.0.1-py3-none-any.whl | Python 3 | none | any | Details for livingcircuit-1.0.1-py3-none-any.whl |
Total release size: 16.6 kB
Release files / livingcircuit-1.0.1.tar.gz
| Download URL | livingcircuit-1.0.1.tar.gz |
|---|---|
| Size | 7.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
d313871d3e2563fa7b264627e69bba03f11219a3aabe2e95a85a02b6e56633ea
|
|
BLAKE2b-256 checksum How to use checksums |
b7cd8a58bcc002b3f18a04024ed1851cd03c848c1d54ce14bf1e5738bc6e655b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.12
|
Release files / livingcircuit-1.0.1-py3-none-any.whl
| Download URL | livingcircuit-1.0.1-py3-none-any.whl |
|---|---|
| Size | 8.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
3913d7a3f8e0bd71309c8d7a0ce782ab3eb42c91b5f5454526d697f39efdd470
|
|
BLAKE2b-256 checksum How to use checksums |
419cbe855de84be1207e6623201df3fae7f0de7ffbffc148b952f57e513f5fb0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.10.12
|
Security for livingcircuit 1.0.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.