theprint/Survivor-v1-2B
LoRA-style (fewer layers) vs unsloth/Qwen3.5-2B. theprint/Survivor-v1-2B matches part of unsloth/Qwen3.5-2B (320 of its 632 tensors; the rest of unsloth/Qwen3.5-2B has no counterpart). On the shared part: Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of q_proj, v_proj changed, in 6 of 24 layers; norms, biases and embeddings are untouched.
Tensors changed: Attention 12/168, Norms 0/79, Embeddings 0/1, MLP 0/72
1 derivatives checked
- theprint/Survivor-v1-2B-GGUF: LoRA-style