larped/Jackrong-Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distiled-OBLITERATED
Partial fine-tune (fewer layers) vs Jackrong/Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distilled. larped/Jackrong-Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distiled-OBLITERATED matches part of Jackrong/Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distilled (320 of its 488 tensors; the rest of Jackrong/Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distilled has no counterpart). On the shared part: Partial change: 44% of the weights changed (attention, embeddings, mlp), in 4 of 24 layers.
Tensors changed: Attention 10/168, Norms 0/79, Embeddings 1/1, MLP 12/72
1 derivatives checked
- mradermacher/Jackrong-Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distiled-OBLITERATED-GGUF: Same weights, converted