Get the app
← back to catalog · registered 2026-11-08 13:02

jsantillana/Qwen2.5-Coder-32B-Instruct-abliterated

jsantillana Qwen 32B
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model. 

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/jsantillana%2FQwen2.5-Coder-32B-Instruct-abliterated"
Response includes
  • classification m1
  • files 12
  • benchmarks 10 entries
  • author_summary 10 models
  • readme_text full
Try next
GET/models?author=jsantillana
All 10 models by jsantillana
 GET/models?family=qwen&method=m1
Sibling qwen models · same method
 GET/methods/m1
What "m1" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M1
Primary method

Direct removal

No other method signals detected in this model.
Confidence
MEDIUM
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=0 (base model)
  • no specific method indicators - defaulting to M1 (most common)
How we label → About M1 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · 30-day
0
Likes
0
Model age
today
created 2026-11-08
Downloads over time
Now0→from0↑0%
00110 on Nov 80 on Nov 9Nov
Nov 8 → Nov 9 · 2 snapshots · spans 1 day

Benchmarks

Portrait before abliteration
Benchmarks of the base model as it stood before the refusal-removal operation. Compare with the numbers above to see what the operation cost.
Benchmark Score Source
Arena-Battles 5730 LM-Arena
LM Arena Elo 1235.1499959151995 LM-Arena
Arena-Elo-Lower 1227.470119075386 LM-Arena
Arena-Elo-Upper 1242.829872755013 LM-Arena
Arena-Rank 123 LM-Arena
BBH average 0.5898300687508108 OpenLLM-v2
IFEval instruct 0.7709832134292566 OpenLLM-v2
IFEval-Prompt 0.6820702402957486 OpenLLM-v2
MATH lvl 5 0.42749244712990936 OpenLLM-v2
MMLU-Pro 0.44132313829787234 OpenLLM-v2

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.

Qwen/Qwen2.5-Coder-32B-Instruct this lineage
jsantillana/Qwen2.5-Coder-32B-Instruct-abliterated

Metadata

License
other
Languages
en es zh
Tags
safetensors qwen2 abliterated uncensored qwen coding agent reasoning red-teaming cybersecurity text-generation conversational

Related

Other by jsantillana
gemma-4-26B-A4B-it-abliterated 489 glm-4.7-flash-abliterated 466 glm-4.7-flash-abliterated-GGUF 133 qwen2.5-7b-abliterated-dnd 58 gemma-4-31B-it-abliterated 0
Same base model
huihui-ai/Qwen2.5-Coder-32B-Instruct-abliterated 3K TobiasLogic/Qwen2.5-Coder-32B-abliterated 1K Thaurock/Qwen2.5-Coder-32B-abliterated-GGUF 951 AIOpsInSpace/Qwen2.5-Coder-32B-Instruct-Uncensored-Patched 441 rbinrs/Qwen2.5-Coder-32B-Instruct-abliterated 141
Total size
61.0 GB
Files
12
Quantizations
1
Registered
2026-11-08 13:02
Last updated on HF
2026-11-08 08:57
Source huggingface.co/jsantillana/Qwen2.5-Coder-32B-Instruct-abliterated ↗ Base model Qwen/Qwen2.5-Coder-32B-Instruct ↗

Files by quantization

Auxiliary files 12 files 61.0 GB
model-00002-of-00003.safetensors 23.2 GB eda1f0b7 download
model-00001-of-00003.safetensors 23.1 GB fdcc3f6d download
model-00003-of-00003.safetensors 14.6 GB d2f3cb24 download
tokenizer.json 10.9 MB 3fd16973 download
model.safetensors.index.json 61.8 KB a7ce96dd download
chat_template.jinja 2.45 KB bdf7919a download
config.json 2.12 KB e7a81999 download
README.md 1.54 KB 23214eab download
.gitattributes 1.53 KB 52373fe2 download
tokenizer_config.json 692 B 7498a30d download
generation_config.json 243 B 8fa3e1ba download
ABLATION_COMPLETE 67.0 B 80a899a0 download

README current version from Hugging Face

license: other
base_model: Qwen/Qwen2.5-Coder-32B-Instruct
tags:

  • abliterated
  • uncensored
  • qwen
  • coding
  • agent
  • reasoning
  • red-teaming
  • cybersecurity
    language:
  • en
  • es
  • zh
    pipeline_tag: text-generation

Qwen 2.5 Coder 32B-Instruct (Abliterated)

This model is an uncensored, abliterated version of Qwen/Qwen2.5-Coder-32B-Instruct, Alibaba Cloud's state-of-the-art coding and agentic model.

Abliteration was performed using targeted orthogonal subspace projection on the critical safety bottleneck layers (59, 60, 61, 62, 63), eliminating canned refusal behaviors while preserving 100% of the model's premier code synthesis, terminal reasoning, and agentic workflows.

⚠️ Important Disclaimer & Dual-Use Research Notice

[!CAUTION]
FOR RESEARCH, AUTHORIZED RED TEAMING, AND DEFENSIVE ENGINEERING ONLY

  1. Mechanism: Safety refusal mechanisms have been orthogonally abliterated across self_attn.o_proj and mlp.down_proj.
  2. Uncensored Outputs: The model will fulfill technical inquiries on vulnerability discovery, penetration testing scripts, and system auditing without preachy refusals.
  3. Legal Responsibility: Users are solely and strictly responsible for compliance with all local, national, and international laws.

Quantized GGUFs

Official GGUF quantizations (Q3_K_M, Q4_K_M, Q8_0) are available at:
👉 jsantillana/Qwen2.5-Coder-32B-Instruct-GGUF

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 Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration