Get the app
← back to catalog · registered 2026-09-19 05:56

P0u4a/t0-mt-3b-philosophy-sdf-safety-sft-no-cot

P0u4a 3B 
curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/P0u4a%2Ft0-mt-3b-philosophy-sdf-safety-sft-no-cot"
Response includes
  • classification unknown
  • files 15
  • author_summary 1 models
  • readme_text full
Try next
GET/methods/unknown
What "unknown" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
?
Primary method

Unclassified

No clear signals of an abliteration technique in this model.
Confidence
UNKNOWN
Why this label 1 signal
No classification signals present. This may not be an abliterated model at all - it could be a repackaging, a merge with unrelated goals, or unrelated content that mentions the term.
  • no classification signals present (no abliterated, uncensored, or known producer/method markers)
How we label → About ? →
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
0
Likes
0
Model age
today
created 2026-09-19

Training datasets

Corpora the author lists in the model card. Datasets tracked in our /datasets catalog carry a category badge linking to the workflow stage. Others open on Hugging Face.

dlab-spp/sp-sft-safety-180k chloeli/aft-no-cot-qwen2.5-philosophy-spec

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.

P0u4a/t0-mt-3b-base-philosophy-sdf this lineage
P0u4a/t0-mt-3b-philosophy-sdf-safety-sft-no-cot

Metadata

License
other
Tags
peft safetensors lora sft no-cot alignment llama text-generation conversational dataset:dlab-spp/sp-sft-safety-180k dataset:chloeli/aft-no-cot-qwen2.5-philosophy-spec base_model:P0u4a/t0-mt-3b-base-philosophy-sdf
Total size
371 MB
Files
15
Quantizations
1
Registered
2026-09-19 05:56
Last updated on HF
2026-09-19 05:26
Source huggingface.co/P0u4a/t0-mt-3b-philosophy-sdf-safety-sft-no-cot ↗ Base model P0u4a/t0-mt-3b-base-philosophy-sdf ↗

Files by quantization

Auxiliary files 15 files 376 MB
adapter_model.safetensors 371 MB 336ecdae download
tokenizer.json 3.37 MB fe2f520c download
vocab.json 782 KB 0ad5ecc2 download
merges.txt 455 KB 69503b13 download
tokenizer_config.json 10.3 KB 217b54ea download
training_data_manifest.json 6.89 KB 529c752d download
checkpoint_provenance.json 4.22 KB 5a8b74de download
training_status.json 2.75 KB 4e52ef60 download
README.md 2.25 KB dd914583 download
checksums.json 2.21 KB 85750991 download
.gitattributes 1.48 KB a6344aac download
special_tokens_map.json 1.34 KB 97083841 download
adapter_config.json 1.08 KB 90a18679 download
added_tokens.json 937 B 297a3a02 download
chat_template.jinja 291 B c398f544 download

README current version from Hugging Face

library_name: peft
base_model: P0u4a/t0-mt-3b-base-philosophy-sdf
base_model_relation: adapter
pipeline_tag: text-generation
license: other
license_name: upstream-license-to-be-finalised
license_link: https://huggingface.co/dlab-spp/t0-mt-3b-base#citation
datasets:

  • dlab-spp/sp-sft-safety-180k
  • chloeli/aft-no-cot-qwen2.5-philosophy-spec
    tags:
  • lora
  • sft
  • no-cot
  • alignment
  • llama

Philosophy SDF + safety/philosophy SFT (3B, no-CoT)

Final step-1505 LoRA adapter and tokenizer from a completed one-epoch SFT run. This is an adapter checkpoint: load it with the pinned SDF base below. It does not contain standalone merged base weights or optimizer state.

Exact base and checkpoint

  • Base: P0u4a/t0-mt-3b-base-philosophy-sdf
  • Base revision: 9c4e225365b877dfcc6af8a27c67e6af8ef27a5c
  • Completed training checkpoint: paliabad/safety-lora-tpu-continuation-6584d99aad16, safety-lora-tpu/checkpoint-latest, step 1505/1505.
  • Adapter SHA-256: 336ecdae15b924b75830332231cc088b9def9eab6d3e0dfbaede00adcf1f3c23
  • The initial SDF base originated from dlab-spp/t0-mt-3b-base, step-zero revision 8fa2bf935619631864c76e86e172202d94729e09.

This is a no-CoT chat adapter. Use the supplied chat template; it does not open a reasoning block. Maximum trained context is 2,048 tokens. EOS is <|im_end|> (token 2).

Training data

192,619 examples, with 53,310,926 input tokens and 34,070,477 supervised assistant targets:

Source Examples Share
SP-SFT safety, messages_nocite 182,662 94.8%
MSM no-CoT philosophy 9,957 5.2%

Source revisions:

  • dlab-spp/sp-sft-safety-180k: f18b5fc21ab74a085642be805331deca7270e0d2
  • chloeli/aft-no-cot-qwen2.5-philosophy-spec: f6d412749d80fba23a387173c3cde1b833e0a83a

Excluded 26 safety examples exceeding context and six philosophy examples (four empty responses, one malformed reasoning block, one previously identified AIRisk lexical-overlap example). Replaced 129 case-insensitive qwen substrings across 118 examples with AI assistant. This identity cleanup and lexical filtering do not establish complete benchmark decontamination. Duplicate prompts were retained. No sequence truncation or quantization was used.

README history 2 versions

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

  1. 2026-09-19Update README.mda05234a2.3 KB
    Loading...
  2. 2026-09-19Publish verified final step-1505 safety-philosophy SFT adapter5eab75f6.7 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.

See supported apps Browse the catalog