Neo
Collection
Neo is a collection of model merges using a novel heterogeneous LLM merging technique (paper in progress). • 7 items • Updated • 1
* evals calculated with llama.cpp llama-perplexity
google/gemma-3-12b-it neopolitized with projected shards and fragments of mistralai/Devstral-2-123B-Instruct-2512.
- projection method: 3
- merge method: 0
- layers: 0-6 [x->x]
- alpha: 0.8-0.8
- tensors: attn_q, attn_k, attn_v
8 w w
8d8b. .d88b .d8b. 88b. .d8b. 8 w w8ww .d88
8P Y8 8.dP' 8' .8 8 8 8' .8 8 8 8 8 8
8 8 `Y88P `Y8P' 88P' `Y8P' 8 8 Y8P `Y88
8
- Downloads last month
- 8
Hardware compatibility
Log In to add your hardware
16-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support