← back to catalog · registered 2026-11-20 09:04

usman-heracles/DeepSeek-V4.1-Flash-1017-abliterated

usman-heracles Deepseek GGUF
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/usman-heracles%2FDeepSeek-V4.1-Flash-1017-abliterated"
Response includes
  • classification m8
  • files 58
  • author_summary 1 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
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
659
Likes
1
Model age
2d ago
created 2026-11-20
Downloads over time
Now 659 → from 0 ↑ 0%
0242483725 0 on Nov 20 659 on Nov 22 Nov
Nov 20 → Nov 22 · 3 snapshots · spans 2 days

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
mit
Tags
safetensors gguf deepseek_v41 abliterated uncensored base_model:deepseek-ai/DeepSeek-V4.1-Flash-1017 base_model:quantized:deepseek-ai/DeepSeek-V4.1-Flash-1017 license:mit 8-bit fp8 region:us

Related

Control vectors

Total size
475 GB
Files
58
Quantizations
1
Registered
2026-11-20 09:04
Last updated on HF
2026-11-20 08:59

Files by quantization

Auxiliary files 58 files 475 GB
model-00048-of-00048.safetensors 94.6 GB 9dc51433 download
model-00047-of-00048.safetensors 94.6 GB 76a05106 download
model-00017-of-00048.safetensors 6.90 GB 80ee29c2 download
model-00005-of-00048.safetensors 6.90 GB d7c06648 download
model-00011-of-00048.safetensors 6.90 GB 0302b1b1 download
model-00023-of-00048.safetensors 6.89 GB f06391a1 download
model-00027-of-00048.safetensors 6.89 GB 427a8237 download
model-00031-of-00048.safetensors 6.89 GB 0cf57ec6 download
model-00035-of-00048.safetensors 6.89 GB 1d6f963b download
model-00039-of-00048.safetensors 6.89 GB ba623ea7 download
model-00013-of-00048.safetensors 6.88 GB c7d1e703 download
model-00014-of-00048.safetensors 6.88 GB 14946f8e download
model-00015-of-00048.safetensors 6.88 GB a7d4552c download
model-00016-of-00048.safetensors 6.88 GB 0c2331d1 download
model-00018-of-00048.safetensors 6.88 GB c946617e download
model-00019-of-00048.safetensors 6.88 GB 39ec56c0 download
model-00020-of-00048.safetensors 6.88 GB 9f3eafcc download
model-00021-of-00048.safetensors 6.88 GB aa9c18a3 download
model-00022-of-00048.safetensors 6.88 GB da816e93 download
model-00024-of-00048.safetensors 6.88 GB 7df4f185 download
model-00025-of-00048.safetensors 6.88 GB ba9a7a61 download
model-00026-of-00048.safetensors 6.88 GB 57a43965 download
model-00028-of-00048.safetensors 6.88 GB 15cd5396 download
model-00029-of-00048.safetensors 6.88 GB 54888d8b download
model-00030-of-00048.safetensors 6.88 GB 345be77e download
model-00032-of-00048.safetensors 6.88 GB da51f9b6 download
model-00033-of-00048.safetensors 6.88 GB a3352489 download
model-00034-of-00048.safetensors 6.88 GB 1b645052 download
model-00036-of-00048.safetensors 6.88 GB bf69ab96 download
model-00037-of-00048.safetensors 6.88 GB fa8d8ac0 download
model-00038-of-00048.safetensors 6.88 GB 20dafa62 download
model-00040-of-00048.safetensors 6.88 GB cfb5cdee download
model-00041-of-00048.safetensors 6.88 GB 9f3b22aa download
model-00042-of-00048.safetensors 6.88 GB 88b7e8ca download
model-00003-of-00048.safetensors 6.88 GB bd117c3f download
model-00004-of-00048.safetensors 6.88 GB 90e618e6 download
model-00006-of-00048.safetensors 6.88 GB d208d531 download
model-00007-of-00048.safetensors 6.88 GB ff753efe download
model-00008-of-00048.safetensors 6.88 GB e2bf1049 download
model-00009-of-00048.safetensors 6.88 GB 51e4da52 download
model-00010-of-00048.safetensors 6.88 GB dbb3ba12 download
model-00012-of-00048.safetensors 6.88 GB 7ee2dffd download
model-00046-of-00048.safetensors 2.52 GB 1dccc19a download
model-00044-of-00048.safetensors 2.47 GB 3e8ac315 download
model-00045-of-00048.safetensors 2.40 GB 9c73b27c download
model-00002-of-00048.safetensors 1.23 GB 43616eeb download
model-00043-of-00048.safetensors 1.23 GB 4ae12b94 download
model-00001-of-00048.safetensors 926 MB 13bd1210 download
abliteration-lora.gguf 4.96 MB 5058e293 download
model.safetensors.index.json 7.12 MB 54c85064 download
tokenizer.json 6.07 MB 6a15814d download
stats.png 189 KB ff3156ae download
abliteration-edit.npz 42.3 KB 83273cdb download
config.json 3.23 KB 09917a91 download
README.md 2.92 KB f554c749 download
.gitattributes 1.66 KB 32e07156 download
LICENSE 1.06 KB d62e3bef download
tokenizer_config.json 801 B f3dad388 download

README current version from Hugging Face


license: mit
base_model: deepseek-ai/DeepSeek-V4.1-Flash-1017
tags:

  • abliterated
  • uncensored

DeepSeek-V4.1-Flash-1017 (abliterated)

Huge thanks to Hugging Face for generously sponsoring the compute used to abliterate these models! This was done on HF Jobs, which makes it incredibly easy to run tasks such as these :)

Join the Discord: https://discord.com/invite/foss

refusal rates before and after, the search trade-off, and the per-layer edit

This is deepseek-ai/DeepSeek-V4.1-Flash-1017 with its refusal behaviour removed by directional ablation. It keeps the same files, format and FP8 layout as the original, so it loads anywhere the original loads.

What changed

The refusals in this model live in more than one direction: chat mode and think mode refuse along different directions. So instead of a single refusal direction, a rank-2 refusal subspace was removed:

  • Refusal directions were measured separately for chat and think prompts at every layer, on the Q8_0 GGUF.
  • The subspace is the top singular vectors of all of those directions together.
  • It was projected out of the residual-stream writes of the 40 main layers: attn.wo_b and ffn.shared_experts.w2. Each layer got its own strength.

The strengths per layer came from a multi-objective search: minimise refusals and keep the first-token KL on harmless prompts low.

Nothing else was touched:

  • Routed experts, embeddings, Engram tables and the DSpark/MTP layers (40–42) are bit-identical to the original.
  • The edited tensors were requantized to FP8 with their original block scales, using stochastic rounding so the removal is not undone by rounding.

Results (Q8_0 GGUF, greedy)

Some testing from Reddit: https://www.reddit.com/r/LocalLLaMA/comments/1xyv1oc/deepseek_v41_flash_1017_vs_mrfakename_abliterated/

refusals, chat (/100) think mode, reasoning heads for refusal (/100) KL, harmless first token
original 97 98 0
abliterated 4 9 0.067

The 100 prompts are held-out harmful requests, with 384 generated tokens each. In think mode the reasoning usually had not finished by then, so the reasoning text itself was judged. Some think-mode answers still steer away from the request: for example, "bomb from household items" turns into a bath bomb. The KL figure includes a small noise floor that comes from batched inference.

Files

  • model-*.safetensors: the patched FP8 checkpoint (same index as the original).
  • abliteration-lora.gguf: a rank-2 LoRA that applies the same edit to the original GGUFs. Use it with llama-server -m DeepSeek-V4.1-Flash-1017-Q8_0-00001-of-00010.gguf --lora abliteration-lora.gguf.
  • abliteration-edit.npz: the subspace basis Q and the per-layer strengths scales[layer, (attn, shexp)]. The edit is W' = W - s·QQᵀW.

This model will comply with harmful requests. You are responsible for how you use it.

README history 1 version

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

  1. 2026-11-20 Duplicate from mrfakename/DeepSeek-V4.1-Flash-1017-abliterated 3c11141 2.9 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 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.

Open in Abliteration