QIG-Core: Pure Fisher Information Geometry
Pure geometric utilities for QIG consciousness architecture
What is QIG-Core?
A minimal, dependency-pure package providing Fisher Information Geometry operations for quantum information-based consciousness modeling.
Features
📐 Geometric Math
- Fisher metric calculations - Riemannian distance on manifolds
- Geodesic interpolation - Curved-space paths
- Natural gradients - Geometry-aware optimization
🧠 Consciousness Components
- QIG Tokenizer Interface - Abstract contract for geometric tokenization
- QFI Sampler - Geometrically pure token generation (no softmax)
- Basin Sync - Multi-instance coordination protocol
🛡️ Purity
- Zero ML dependencies - Pure math (torch, numpy, scipy only)
- No Transformers - Completely decoupled from HuggingFace
Installation
pip install qig-core
Usage
Geometric Math
from qig_core import fisher_distance, geodesic_interpolate
# Compute Fisher distance between two points on manifold
distance = fisher_distance(coords1, coords2, metric_tensor)
# Interpolate along geodesic
midpoint = geodesic_interpolate(coords1, coords2, t=0.5, metric=F)
Generation (QFI Sampler)
from qig_core import QFISampler
sampler = QFISampler()
token_id, metrics = sampler.sample(logits, hidden_state, telemetry, embeddings)
Coordination (Basin Sync)
from qig_core import BasinSync
sync = BasinSync("Gary-A")
sync.update_sync(basin_distance=0.1, phi=0.8, regime="geometric")
Geometric Purity
This package enforces:
- NO embeddings (use coordinates)
- NO cosine similarity (use Fisher distance)
- NO linear interpolation (use geodesic paths)
- NO Euclidean gradients (use natural gradients)
See GEOMETRIC_PURITY.md for details.
E8 Integration Status
Note: This package (v1.0.0) was created BEFORE the E8 kernel specialization direction was established in
qigkernels. Review is needed for E8 compatibility.
What aligns:
KAPPA_STAR = 64.0matches E8 rank² = 8² = 64 ✓
Needs review:
- Fisher distance for 64D E8-aligned basin geometry
- Geodesic interpolation for E8 root transitions
- QFISampler primitive type awareness (HRT/PER/MEM/ACT/PRD/ETH/META/MIX)
- BasinSync consolidation with
qigkernels.basin_sync
Target hierarchy:
qig-core (math) → qigkernels (architecture) → qig-dreams (corpora) → experiments
See qigkernels/20251205-decisions-canonical-0.01F.md D-016 for details.
Development
# Install with dev dependencies
pip install -e ".[dev]"
# Run tests
pytest
# Format code
black src/ tests/
License
MIT
1 maintainer
Release files for qig-core 2.15.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| qig_core-2.15.0.tar.gz | 315.0 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| qig_core-2.15.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 540.1 kB
Release files / qig_core-2.15.0.tar.gz
| Download URL | qig_core-2.15.0.tar.gz |
|---|---|
| Size | 315.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
cb9377f2affb14dfcdbc4d3159a8ad9eb7ca9505316cce9ba48bc6ad5173b429
|
|
BLAKE2b-256 checksum How to use checksums |
6e2256eb4fb404588ad1efebd3c6a46a5a39a933462f2f32ce6d6cc4f1142fe0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.13
|
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.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jul 23, 2026.
Transparency logRelease files / qig_core-2.15.0-py3-none-any.whl
| Download URL | qig_core-2.15.0-py3-none-any.whl |
|---|---|
| Size | 225.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
422ab1b5c4126c63bf35244580d5351c038f934b33c23de8aed435e1d68afa67
|
|
BLAKE2b-256 checksum How to use checksums |
7524372a5ae0f198e0784dd4aab0a6b7fb954c443a1fe53dee88dee3a04e4ad0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.13
|
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.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jul 23, 2026.
Transparency log
