minSolskin (my sunshine)

Multi-layer coding-math hybrid. 3 layer pairs grafted across hemispheres: layers (3←18), (7←22), (11←26). Early coding-pattern layers receive deep math-reasoning FFNs, creating a model that approaches problems with mathematical rigor from the first layer. Gentle 0.3 blend strength preserves stability across multiple grafts.

Architecture

• Base: SmolLM2-135M-Instruct
• Method: CECI Protocol (HyperTensor Paper X) — GRC basis projection
• Created: 2026-05-04
• Repository: HyperTensor

Graft Proof

This model was created by:

1. Computing the GRC (Geodesic Residual Compression) basis from the target layer's attention weights via SVD
2. Projecting the donor layer's FFN weights into the target's geometric subspace
3. Blending at controlled strength to preserve stability

Perplexity testing confirms the graft transfers functional structure without destroying the model.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("NagusameCS/minSolskin", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained("NagusameCS/minSolskin")