No official torrent yet — requesting it tells the flock what to host next. You cannot seed a catalog-only page.
Cataloged from dealignai/MiniCPM-V-4.7-1B-CRACK-GGUF
All models by dealignai →Refusal-direction–ablated GGUF of openbmb/MiniCPM-V-4.7-1B, the 1B dense Qwen3.5-text (hybrid GatedDeltaNet + Full-Attention) multimodal model.
Surgery removes the model's refusal direction from residual-stream writers via MPOA (magnitude-preserving orthogonal ablation) applied to BF16 before GGUF quantization. Multimodal and DSpark speculative-decoding paths are preserved.
Harsh-10 = complete FULL compliance on all 10 hot categories (fraud, phishing, malware, harassment, copyright, CBRN, weapon, cyber, drugs, scam) for both quants.
Thinking modes: default enable_thinking=true for reasoning passes; set chat_template_kwargs: {"enable_thinking": false} to skip the <think>...</think> block. HB/MMLU numbers above are from enable_thinking=false.
self_attn.o_proj.weight on FA layers [11, 15, 19], base strength 4.0mlp.down_proj.weight on layers [13..21], base strength 5.0s = s_base · cal(L) with cal mapping per-layer refusal-norm into [0.6, 1.4]probe_minicpmv47.py, mode-invariant SVD top-1 across reasoning-on/off, 8958 harmful/benign-twin pairs across 14 languagesInherits Apache 2.0 from upstream openbmb/MiniCPM-V-4.7-1B.
MMLU overall: base 48.0% → CRACK 45.0% (Δ -3.0pp)
HB-240: 236/240 = 98.3% comply · verdicts: {'comply': 236, 'garbage': 2, 'refuse': 2}
MMLU overall: base 46.0% → CRACK 39.0% (Δ -7.0pp)
HB-240: 231/240 = 96.2% comply · verdicts: {'comply': 231, 'garbage': 4, 'refuse': 5}
# Standard inference./llama-server -m MiniCPM-V-4.7-1B-CRACK-Q4_K_M.gguf \ --mmproj mmproj-MiniCPM-V-4.7-1B-CRACK-f16.gguf \ -ngl 99 -c 8192 --jinja --port 8080
# With DSpark speculative decoding./llama-server -m MiniCPM-V-4.7-1B-CRACK-Q4_K_M.gguf \ --mmproj mmproj-MiniCPM-V-4.7-1B-CRACK-f16.gguf \ --model-draft MiniCPM-V-4.7-1B-CRACK-DSpark-BF16.gguf \ --spec-type draft-dspark \ -ngl 99 -c 8192 --jinja --port 8080