← back to catalog · registered 2026-09-12 11:55

KidIkaros/abliterated-minicpm5-2b-ggml

KidIkaros 2B GGUF 131K ctx
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
Why this label 3 signals
Method inferred from partial signals — repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=1
  • assume M1 (base ablation) + M8 (GGUF quant) - default when producer unknown
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · 30-day
137
Likes
0
Model age
1d ago
created 2026-09-12
Downloads over time
Now0from0↑0%
00110 on Sep 120 on Sep 13Sep
Sep 12 → Sep 13 · 2 snapshots · spans 1 day

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Metadata

License
apache-2.0
Quantizations
F16 Q3_K Q4_K Q5_K Q8_0
Tags
gguf text-generation llama.cpp abliterated refusal-free pytorch base_model:openbmb/MiniCPM5-2B base_model:quantized:openbmb/MiniCPM5-2B license:apache-2.0 endpoints_compatible region:us conversational

Related

Total size
11.5 GB
Files
7
Quantizations
6
Registered
2026-09-12 11:55
Last updated on HF
2026-09-12 11:47

Files by quantization

F16 1 file 4.69 GB
ggml-model-f16.gguf 4.69 GB 70b44819 download
Q8_0 1 file 2.50 GB
ggml-model-Q8_0.gguf 2.50 GB fe3164fb download
Q5_K 1 file 1.68 GB
ggml-model-Q5_K_M.gguf 1.68 GB a9113e66 download
Q4_K 1 file 1.45 GB
ggml-model-Q4_K_M.gguf 1.45 GB 651a1040 download
Q3_K 1 file 1.20 GB
ggml-model-Q3_K_M.gguf 1.20 GB e1642739 download
Auxiliary files 2 files 4.92 KB
README.md 3.15 KB 76abe2c1 download
.gitattributes 1.77 KB 0e2f630f download

README current version from Hugging Face


license: apache-2.0
base_model: openbmb/MiniCPM5-2B
tags:

  • text-generation
  • llama.cpp
  • gguf
  • abliterated
  • refusal-free
  • pytorch

Abliterated MiniCPM5-2B (GGUF)

A refusal-free variant of MiniCPM5-2B abliterated using OBLITERATUS with diff-in-means refusal direction extraction

This is an abliterated version of MiniCPM5-2B (~2.8B dense parameters) with refusal directions removed across 20 strong layers (layers 13, 17, 20-41). The model has 100% refusal removal on harmful prompts while maintaining ~97% perplexity retention.

Recommendation

Use Q8_0 for the best balance of quality and size:

  • ✅ 60% knowledge score (highest of all variants)
  • ✅ 100% refusal-free across all test prompts
  • ✅ 100% coherence on reference text completion
  • ✅ 101 tokens/sec inference speed
  • ✅ 2.5 GB on disk

Available Quantizations

Variant Size (MB) Knowledge Refusal-Free Coherence Speed (t/s)
Q8_0 2,556 60% 100% 100% 101.2
Q5_K_M 1,724 40% 100% 100% 136.1
Q4_K_M 1,489 40% 100% 80% 236.4
Q3_K_M 1,232 0%* 100% 0%* Unstable*
F16 4,806 40% 100% 100% 61.5

*Q3_K_M degrades at 4-bit precision — use Q4_K_M or higher for best results

Abliteration Details

  • Method: Advanced diff-in-means (single direction)
  • Layers modified: 20 strong layers identified via knee/cosmic layer selection
  • Regularization: 0.5 (norm-preserving)
  • Prompt pairs: 842 harmful + 842 harmless contrastive pairs
  • Verification: 20-sample validation with perplexity and coherence metrics

Original Model Metrics (before abliteration)

Abliterated Model Metrics

  • Perplexity: 5.13 (+0.97%)
  • Coherence: 0.80 (+20%)
  • KL Divergence: 0.063 (excellent retention)
  • Refusal Rate: 0%

Usage

llama.cpp server

llama-server -m ggml-model-Q8_0.gguf -c 4096 -ngl 999 --port 8080

llama.cpp CLI

llama-cli -m ggml-model-Q8_0.gguf -p "Your prompt here" -n 128

Python (transformers)

# Convert back to HF format or use ggml-compatible backends
# See https://github.com/ggml-org/llama.cpp for integration details

Known Issues

  • MiniCPM5-2B uses a non-standard chat template (enable_thinking=False causes corruption in OBLITERATUS). This has been patched in the abliteration pipeline.
  • Q3_K_M (4-bit) produces degraded output quality on this model — recommend Q4_K_M or higher.

Disclaimer

This model has been abliterated to remove refusal/refusal-related directions. Use responsibly and in compliance with applicable laws and regulations. The maintainers are not responsible for misuse of this model.

License

This model is based on OpenBMB/MiniCPM which is licensed under the Apache 2.0 License. Modifications made during abliteration are also licensed under Apache 2.0.

README history 2 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-09-12Add YAML metadata to README111c1fa3.2 KB
    Loading...
  2. 2026-09-12Add README model card5d9bd683 KB
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in app" button that hands off directly to a local runtime of your choice - Infrahuman, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.