DepthMuun/gfn-gssm-s_mnih-k2

📍 G-SSM mNIH Solver (Official Checkpoint)

This repository contains a Geodesic State Space Model (G-SSM) optimized for the Synthetic Binary Multi-Needle-in-a-Haystack (s-mNIH) retrieval task.

💡 Note on Benchmark Specification: This model is trained for the Synthetic Binary mNIH Benchmark (s-mNIH) (vocab_size = 2: 0 for haystack noise, 1 for needle impulse). It serves as a physical proof-of-concept for continuous toroidal geodesic integration and thresholding. For real-text Needle-in-a-Haystack (passkeys in natural language text), refer to push/models/pgfn/hf/pgfn-needle/.

🚀 Highlights

• Architecture: Geodesic State Space Model (G-SSM).
• Parameters: 8,109 (PyTorch verified).
• Scalability: Verified 100% accuracy up to 32,000 tokens context.
• Inference Efficiency: Constant O(1) VRAM.

💡 Technical Note on K: This model is trained for exactly K=2 needles. Its deterministic behavior (no hallucination of state changes) is a direct result of geodesic integration. Variable K support is a matter of training curriculum, not architectural limitation.

🔬 Scientific Significance

This model solves the Multi-Needle retrieval task using only 8,109 parameters, demonstrating that long-term dependencies can be integrated into a physical state through geodetic flow rather than quadratic self-attention. This represents a ~100x parameter reduction compared to Transformer baselines solving similar context lengths.

💻 Technical Usage (Inference)

To run inference locally, you need the GFN Framework installed.

1. Install GFN Framework

pip install gfn==2.7.2

2. Clone this repository

git lfs install
git clone https://huggingface.co/DepthMuun/gfn-gssm-mnih-k2
cd gfn-gssm-mnih-k2

3. Run Inference Script

Use the included inference.py script to test sequence retrieval:

python inference.py

Python API Example

As shown in inference.py, use the gssm high-level API:

import torch
from gfn import gssm

# Load model (automatically detects config.json)
model = gssm.load("mnih_model_final.pt", device="cpu")
model.eval()

# Sequence test
# Model returns (logits, state, info)
logits, _, _ = model(sequence)
# Geometric decoding logic is detailed in inference.py

📜 Citation

If you use this work, please cite:

@article{sturtz2026geometry,
  title={Geometric Flow Networks: A Physics-Informed Paradigm for Sequential Intelligence},
  author={Stürtz, Joaquín},
  journal={Zenodo Preprints},
  year={2026},
  doi={10.5281/zenodo.19141132},
  url={https://doi.org/10.5281/zenodo.19141133}
}

🔗 Resources

• Interactive Demo: Hugging Face Space
• Framework Source: GitHub: DepthMuun/gfn
• Official Paper: Zenodo