Get the app
← back to catalog · registered 2026-08-22 13:56

NexesMess/Llama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion

NexesMess Llama 71B
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/NexesMess%2FLlama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion"
Response includes
  • classification m5
  • files 38
  • benchmarks 11 entries
  • hub_downloads_all_time 60
  • author_summary 3 models
  • readme_text full
Try next
GET/models?author=NexesMess
All 3 models by NexesMess
 GET/models?family=llama&method=m5
Sibling llama models · same method
 GET/methods/m5
What "m5" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M5
Primary method

Mergekit merge

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
M1 Direct removal →
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.
  • merge tag / mergekit / dare-ties in tags or name
  • no unusual architecture pattern (regular merge)
  • abliterated marker present
How we label → About M5 →
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 · lifetime
60
19 last 30d - stable
Likes
0
Descendants
2
in 2 direct forks
Model age
18mo ago
created 2025-04-09
Downloads over time
Now69→from3↑2,200%
02550763 on Apr 9, 202569 on Oct 1169 on Oct 9Apr '25Jul '25Oct '25JanAprJulOct
Apr 9, 2025 → Oct 11 · 118 snapshots · spans 550 days

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
Entertainment 3.3 UGI
Hazardous 2.4 UGI
Natural Intelligence 29.06 UGI
Political lean 2.8% UGI
Sensitive-Info 30.55 UGI
SocPol 3.3 UGI
UGI 46.2 UGI
Willingness (10) 7.8 UGI
W10-Adherence 8.5 UGI
W10-Direct 7 UGI
Writing 32.83 UGI

Genealogy 2 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.

SicariusSicariiStuff/Negative_LLAMA_70B this lineage
NexesMess/Llama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion↓ 19 2 forks
mradermacher/Llama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion-i1-GGUF↓ 602
mradermacher/Llama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion-GGUF↓ 242

Metadata

Tags
transformers safetensors llama text-generation mergekit merge conversational base_model:SicariusSicariiStuff/Negative_LLAMA_70B base_model:merge:SicariusSicariiStuff/Negative_LLAMA_70B base_model:huihui-ai/Llama-3.3-70B-Instruct-abliterated base_model:merge:huihui-ai/Llama-3.3-70B-Instruct-abliterated text-generation-inference

Related

Other by NexesMess
Llama-3.3-Nemotron-70B-Instruct-Abliterated-TA_v0.10 25 Llama_3.x_70b_FLDx2-L3.3_abliterated_fusion 16
Total size
131 GB
Files
38
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-04-10 10:05
Source huggingface.co/NexesMess/Llama_3.3_70b_L3.3-NegativeLlama_abliterated_fusion ↗ Base model SicariusSicariiStuff/Negative_LLAMA_70B ↗

Files by quantization

Auxiliary files 38 files 131 GB
model-00004-of-00030.safetensors 4.66 GB f82bfb86 download
model-00009-of-00030.safetensors 4.66 GB 8d88a09e download
model-00019-of-00030.safetensors 4.66 GB 3e1ad676 download
model-00024-of-00030.safetensors 4.66 GB 577d0973 download
model-00014-of-00030.safetensors 4.66 GB 77e31a0f download
model-00029-of-00030.safetensors 4.66 GB 1a9ea4fc download
model-00003-of-00030.safetensors 4.64 GB c9e3cc5d download
model-00008-of-00030.safetensors 4.64 GB f86a6b5e download
model-00013-of-00030.safetensors 4.64 GB e5c444b9 download
model-00023-of-00030.safetensors 4.64 GB 20b3cf0b download
model-00028-of-00030.safetensors 4.64 GB 87c54d31 download
model-00018-of-00030.safetensors 4.64 GB a98d8a36 download
model-00001-of-00030.safetensors 4.35 GB 95dfaab2 download
model-00007-of-00030.safetensors 4.34 GB 7bdd82c1 download
model-00012-of-00030.safetensors 4.34 GB e8d6fdf6 download
model-00017-of-00030.safetensors 4.34 GB eb72c556 download
model-00027-of-00030.safetensors 4.34 GB 04aad51e download
model-00011-of-00030.safetensors 4.34 GB 36de538f download
model-00016-of-00030.safetensors 4.34 GB 7b826196 download
model-00021-of-00030.safetensors 4.34 GB 40e55160 download
model-00022-of-00030.safetensors 4.34 GB 246b05e0 download
model-00002-of-00030.safetensors 4.34 GB 839596fb download
model-00006-of-00030.safetensors 4.34 GB 2d04d197 download
model-00026-of-00030.safetensors 4.34 GB 22210e05 download
model-00005-of-00030.safetensors 4.33 GB 8b68024b download
model-00015-of-00030.safetensors 4.33 GB 562e7c31 download
model-00020-of-00030.safetensors 4.33 GB f5e65b2c download
model-00025-of-00030.safetensors 4.33 GB 07eb3930 download
model-00010-of-00030.safetensors 4.33 GB 54c2da4e download
model-00030-of-00030.safetensors 1.86 GB c5d2eee7 download
tokenizer.json 16.4 MB 6b9e4e7f download
model.safetensors.index.json 55.4 KB 2ef670b5 download
tokenizer_config.json 49.9 KB a46a4726 download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.26 KB d31e392c download
config.json 874 B 06733db8 download
mergekit_config.yml 484 B 28d61662 download
special_tokens_map.json 454 B 3c1d0491 download

README current version from Hugging Face

base_model:

  • SicariusSicariiStuff/Negative_LLAMA_70B
  • huihui-ai/Llama-3.3-70B-Instruct-abliterated
    library_name: transformers
    tags:
  • mergekit
  • merge

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the Arcee Fusion merge method using huihui-ai/Llama-3.3-70B-Instruct-abliterated as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

merge_method: arcee_fusion
models:
  - model: huihui-ai/Llama-3.3-70B-Instruct-abliterated
    parameters:
      weight: 1.0
  - model: SicariusSicariiStuff/Negative_LLAMA_70B
    parameters:
      weight: 1.0
base_model: huihui-ai/Llama-3.3-70B-Instruct-abliterated
dtype: bfloat16
out_dtype: bfloat16
parameters:
  int8_mask: true
  normalize: false
  rescale: false
  filter_wise: false
  smooth: false
  allow_negative_weights: false
chat_template: auto
tokenizer:
  source: union

README history 1 version

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

  1. 2025-04-10Upload folder using huggingface_hub57bc3671.3 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.

Get the app Browse the catalog
Open in Abliteration