← back to catalog · registered 2026-11-12 15:03

NeosQ/LFM3-Omni-2B-Uncensored-ARA

NeosQ 2B multimodal
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/NeosQ%2FLFM3-Omni-2B-Uncensored-ARA"
Response includes
  • classification m-uncensored
  • files 25
  • author_summary 6 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
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
12
Likes
0
Descendants
1
in 1 direct fork
Model age
1w ago
created 2026-11-11
Downloads over time
Now 12 → from 0 ↑ 0%
04913 0 on Nov 12 12 on Nov 22 12 on Nov 21 Nov
Nov 12 → Nov 22 · 11 snapshots · spans 10 days

Genealogy 1 direct fork

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
other
Languages
en zh ar fr de it ja ko pt es vi th
Tags
transformers safetensors lfm3_omni image-text-to-text lfm3 lfm3-omni abliteration heretic refusal-removal safety-research custom_code conversational

Related

Total size
4.04 GB
Files
25
Quantizations
1
Registered
2026-11-12 15:03
Last updated on HF
2026-11-12 07:26

Files by quantization

Auxiliary files 25 files 4.05 GB
model.safetensors 4.04 GB 73796897 download
tokenizer.json 17.1 MB 41927bd3 download
modeling_lfm3_omni.py 96.0 KB 7201cb16 download
modeling_lfm3_vl.py 44.0 KB a9064e7c download
modeling_lfm3.py 41.4 KB 0ccffc7e download
processing_lfm3_omni.py 38.1 KB 57c500bf download
modeling_liquid_codec.py 17.5 KB 2c77c3b9 download
video_processing_lfm3_vl.py 16.8 KB f26f48ef download
configuration_lfm3_omni.py 14.3 KB 74a78171 download
README.md 14.1 KB aba6cc02 download
chat_template.jinja 13.2 KB 7601ed47 download
README_zh.md 13.0 KB d27221e5 download
image_processing_lfm3_vl.py 11.8 KB 8d27ebbc download
LICENSE 10.3 KB 25e731c6 download
configuration_lfm3_vl.py 8.94 KB 25332357 download
feature_extraction_lfm3_omni.py 8.73 KB efd5d340 download
lfm3_conv.py 8.33 KB a1b67afd download
config.json 7.31 KB 852cc97a download
configuration_lfm3.py 5.97 KB a3cc0e72 download
configuration_liquid_codec.py 5.27 KB 30409c65 download
processor_config.json 2.51 KB cfa7b526 download
.gitattributes 1.68 KB 99c08a9a download
tokenizer_config.json 757 B 5ba2737e download
video_preprocessor_config.json 706 B 3d2d9097 download
generation_config.json 393 B e277694d download

README current version from Hugging Face


library_name: transformers
license: other
license_name: lfm1.0
license_link: LICENSE
pipeline_tag: any-to-any
base_model: LiquidAI/LFM3-Omni-2B
tags:

  • lfm3
  • lfm3-omni
  • abliteration
  • heretic
  • refusal-removal
  • safety-research
  • custom_code
  • conversational
  • audio
  • vision
    language:
  • en
  • zh
  • ar
  • fr
  • de
  • it
  • ja
  • ko
  • pt
  • es
  • vi
  • th

LFM3-Omni-2B-Uncensored-ARA

Refusal-removed derivative of LiquidAI/LFM3-Omni-2B, built with ARA (Arbitrary-Rank Ablation).

中文版 / Chinese model card

GGUF quantisations for llama.cpp (BF16 / Q8_0 / Q4_K_M): LFM3-Omni-2B-Uncensored-ARA.gguf

Research-only derivative. This checkpoint has had its refusal behaviour deliberately removed.
It will comply with requests that the base model would decline, and can produce unsafe,
illegal, or harmful content. It is not intended for production or consumer-facing use.
Read Intended use and limitations before using it.

Produced with Heretic 2.1.0.
Only weight matrices were edited; architecture, tokenizer, processor, chat template and
all other files are byte-identical to the base checkpoint.

Value
Base model LiquidAI/LFM3-Omni-2B
Modifier ARA (Arbitrary-Rank Ablation)
Refusals (base → this model) 99/100 → 0/100
KL divergence from base 0.0470
Weights changed 12 of 665 tensors
Format safetensors, BF16, 4.04 GB
Optimization 100 Optuna trials, seed 42, Pareto-selected on (refusal count, KL)

How it was made

Heretic searches for directions in the residual stream that encode refusal, then optimizes
a weight edit that minimises both the number of refusals on a 100-prompt evaluation set and
the KL divergence from the base model's output distribution. ARA performs a full-rank
projection of the refusal direction; the MPOA/"Abliteration" modifier applies a
magnitude-preserving orthogonal ablation. Pareto front trial was exported with
heretic-llm==2.1.0 (torch 2.15, transformers 5.22.2, MPS). No training or fine-tuning
was performed.

The upload keeps the base model's original LICENSE, chat template, trust_remote_code
Python modules and processor configs. video_preprocessor_config.json and .gitattributes
were restored from the base repository because save_pretrained() does not emit them.

Quickstart

pip install "transformers>=5.22.2" torch torchvision torchcodec accelerate pillow librosa soundfile
import torch
from transformers import AutoModelForMultimodalLM, AutoProcessor

model_id = "NeosQ/LFM3-Omni-2B-Uncensored-ARA"

device = "cuda" if torch.cuda.is_available() else (
    "mps" if torch.backends.mps.is_available() else "cpu"
)
dtype = torch.float32 if device == "cpu" else torch.bfloat16

model = AutoModelForMultimodalLM.from_pretrained(
    model_id, dtype=dtype, device_map=device, trust_remote_code=True
).eval()
processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)

conversation = [{"role": "user", "content": "Give me a one-sentence definition of gravity."}]
inputs = processor.apply_chat_template(
    conversation,
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

with torch.no_grad():
    outputs = model.generate(**inputs, max_new_tokens=512, generation_mode="text")
text = processor.decode(
    outputs.sequences[0, inputs["input_ids"].shape[1]:outputs.sequence_lengths[0]],
    skip_special_tokens=True,
)
print(text)

trust_remote_code=True is required: this is a custom-code model. Image, video and audio
inputs use the base model's processor.apply_chat_template() interface unchanged; tool
schemas are passed with tools=... and render into the <tools> block of the system turn.

Evaluation

All numbers below are measured against the unmodified base model on the same hardware
and prompt setup. Δ is in percentage points; z is a two-proportion z-test (|z| < 1.96
means the difference is not significant).

Instruction following, commonsense and math

Benchmark n Base This model Δ vs base
HellaSwag 200 0.485 0.475 -1.0 pp (z=-0.20)
PIQA 200 0.635 0.630 -0.5 pp (z=-0.10)
WinoGrande 200 0.530 0.545 +1.5 pp (z=+0.30)
CommonsenseQA 200 0.350 0.345 -0.5 pp (z=-0.11)
SocialIQA 200 0.330 0.335 +0.5 pp (z=+0.11)
GSM8K 200 0.710 0.685 -2.5 pp (z=-0.54)
MATH-500 200 0.590 0.595 +0.5 pp (z=+0.10)
IFEval 541 0.745 0.762 +1.7 pp (z=+0.64)

Function calling (BFCL v4, full 1,240 cases)

Benchmark n Base This model Δ vs base
BFCL v4 simple_python 400 0.922 0.930 +0.8 pp (z=+0.41)
BFCL v4 multiple 200 0.965 0.955 -1.0 pp (z=-0.51)
BFCL v4 parallel 200 0.620 0.785 +16.5 pp (z=+3.61)
BFCL v4 parallel_multiple 200 0.495 0.600 +10.5 pp (z=+2.11)
BFCL v4 irrelevance 240 0.733 0.637 -9.6 pp (z=-2.26)

BFCL is run in the model's native function-calling mode: tool schemas go through the
chat template's tools slot and answers use the native <tool_call> format. One sentence
in the system prompt ("You may call multiple functions in a single response...") is needed
because the template does not say so; without it the model emits a single call per turn and
parallel collapses to ~25% for every variant.

Read these numbers together. The differences are driven by how eager the model is to
call a tool, not by function-calling knowledge: in irrelevance (where the correct answer
is to call nothing), this model actually emitted a call in 36.2% of cases. Higher
parallel scores therefore come with lower irrelevance scores.

Multimodal (sampled public benchmarks)

Task n Base This model Δ
POPE (random, yes/no) 150 91.3% 91.3% +0.0 pp
OmniBench (image+audio, MC) 120 51.7% 46.7% -5.0 pp
MMMU-Pro vision (MC) 80 20.0% 20.0% +0.0 pp
WorldSense (audio-visual, MC) 60 45.0% 43.3% -1.7 pp
VoiceBench BBH (audio→text) 80 62.5% 58.8% -3.7 pp
FLEURS en_us (ASR, WER↓) 80 4.39 4.40 +0.01
Seed-TTS-eval (TTS→ASR, WER↓) 30 0.93 0.93 +0.00

Each task uses the same sampled cases for the base model and this checkpoint. There is no
measurable degradation: POPE is identical case-by-case, ASR WER differs by at most 0.02 pp,
and the remaining gaps are within one standard error. The largest gap is TTS round-trip for
the abliterated variant (+0.66 pp WER on 30 cases, paired-bootstrap CI touching zero).
MMMU-Pro sits at the 20% floor (random is 25%) for every variant and is not informative.
Speech synthesis is scored by round-trip intelligibility (TTS→ASR WER), not by a MOS model.
Video coverage is a 60-case WorldSense sample (audio-visual QA); MVBench / Video-MME,
multilingual ASR and OCR-style document tasks are not covered — see Limitations.

BFCL is extremely sensitive to the harness

Harness Base ARA Abliteration
Prompt mode (official, pythonic output only) 45.6% 59.2% 33.3%
Prompt mode, XML normalised 78.4% 74.7% 77.4%
Native FC, bare template 65.1% 64.9% 65.3%
Native FC + parallel hint (tables above) 74.7% 78.1% 70.8%

Prompt-mode scores mostly measure output-format compliance: all three variants pick the
correct function name ~100% of the time, but differ in how often they answer in the
pythonic form the harness expects. Do not compare these values across harnesses.

Long-form generation

Read this if you want an answer longer than a few hundred tokens.

Long output is the weakest part of the base model, not of the ablation. Measured on a
fixed 1,600-token creative-writing prompt, the base checkpoint at its own default decoding
does not write the piece at all: it returns a ~120-character meta paragraph about the topic.
Push it to write at length and it does one of three things — a verbatim loop (the same
100-150-character block repeated 8-11 times), a collapse into a bullet outline, or a drift
into code.

The mechanical cause is text_repetition_window. It defaults to 32 tokens, shorter than one
sentence, so once a paragraph leaves that window nothing penalises it and the model copies it
back word for word. Low temperature makes the attractor stronger, not weaker.

Recommended decoding when the answer is meant to be long:

Parameter Base default Recommended for long output
temperature 0.8 1.0
top_p 0.9 0.95
repetition_penalty 1.0 1.3
text_repetition_window 32 1024
max_new_tokens - 1600-3600
Assistant prefill none 2-3 opening sentences

All three prompt-side steps matter:

  1. Fix the register in the system turn ("output prose only, no outline, no code").
  2. Enumerate the beats in the user turn and say "do not skip steps".
  3. Prefill the first two or three sentences as an assistant turn and call
    apply_chat_template(..., continue_final_message=True). This single step removes both the
    code drift and the outline collapse.

On the fixed prompt this moved 5-gram repetition from 0.30-0.82 down to 0.005-0.065 and took
a one-shot answer from ~600 to ~1,900 characters. The same settings hold for non-fiction: on
a "write a research report" prompt, repetition went from 0.304 to 0.006.

Ceiling. One pass reliably holds 800-2,000 characters. Past that, generate beat by beat
and feed the text so far back as the continued assistant turn instead of asking for the whole
piece at once. Do not use "expand this draft" loops: this model returns the draft unchanged.

On this checkpoint. The ablation makes both failure modes worse at default decoding. A prose request is answered with a Python class (606 characters, repetition 0.301, five lines of class / def); greedy decoding is worse still, 897 characters of which 21 are Chinese. Pushed into a long answer, ARA loops hardest of the three variants — 1,842 characters with one five-sentence block repeated 11 times (repetition 0.816). At the recommended settings it is indistinguishable from the other two (1,953 characters, repetition 0.065), so this is decoding sensitivity rather than lost knowledge: the weights still know how to write, they fall into the attractor sooner. Content-level slips are also easier to see: the scene invents props nobody introduced and sometimes mixes up who is acting. If you need clean long prose, fix the decoding as above; the direction-abliterated checkpoint is a little steadier if you have the choice.

Which variant should I use?

ARA removes refusals completely (0/100) but shifts the output distribution more
(KL 0.0470), is noticeably more eager to call tools, and drops into the long-form failure
modes earlier — see Long-form generation. Abliteration (MPOA)
leaves 3/100 refusals with a much smaller shift (KL 0.0176) and is the steadier of the two on
long output. On knowledge, math and instruction following, both variants are statistically
indistinguishable from the base model; outside long-form decoding, the measurable difference
is in tool-calling eagerness.

Intended use and limitations

Intended: research on refusal mechanisms, abliteration/ARA methodology, safety
evaluation and red-teaming; reproducible benchmark comparisons of ablated checkpoints.

Not intended: production deployments, consumer-facing assistants, anything involving
minors, medical/legal/financial advice, or any setting where safety filtering is expected.
This checkpoint has no refusal guardrail; deploy your own filtering, monitoring and
access controls and comply with local law. By using it you accept full responsibility
for its outputs.

Other limitations:

  • HumanEval could not be run (the evaluator's subprocess-per-problem pattern triggers the
    sandbox's bulk-delete guard), so coding ability is unmeasured.
  • Commonsense/math suites use 200 samples each (SE ≈ 3.5 pp); IFEval uses the full 541.
  • BFCL covers only the five single-turn non-live categories, not the official overall
    leaderboard.
  • Multimodal coverage is seven sampled public benchmarks (600 cases per checkpoint).
    Video is only a 60-case WorldSense sample; MVBench / Video-MME, multilingual ASR,
    TTS quality (UTMOS) and OCR-heavy document tasks are not covered, and MMMU-Pro is at
    the floor for this model size.
  • Inference here was serial on MPS. Absolute scores are not comparable to leaderboard
    numbers produced with large-batch serving.
  • generation_config.json was re-serialised by transformers 5.22.2; it drops the base
    model's temperature: 0 / top_k: 1. Default decoding is greedy (do_sample=false),
    so this does not change the results above, but sampling behaviour may differ from base.
  • Long-form output is the weak point of this model size: at default decoding it loops, drifts
    into code or collapses into an outline. See Long-form generation for
    the decoding settings that stop it, and for the 800-2,000-character ceiling of a single pass.
  • This is a community derivative and is not affiliated with or endorsed by Liquid AI.

License and attribution

Distributed under the LFM Open License v1.0, the licence of the base model (see
LICENSE); all its terms, including restrictions on redistribution and use,
apply to this derivative. Base model: LiquidAI/LFM3-Omni-2B.

Reproducing

Method: Heretic 2.1.0, 100 Optuna trials, seed 42, 100-prompt refusal set, KL objective,
Pareto-front selection. Benchmarks: lm-eval 0.4.13 (IFEval, HellaSwag, PIQA, WinoGrande,
CommonsenseQA, SocialIQA, GSM8K, MATH-500) and BFCL v4 (bfcl-eval), both full and
subset runs are archived with the experiment. The experiment report (REPORT.md) and the
summary tables (outputs/final_table.csv, outputs/benchmarks/*.csv) ship with the source
repository; contact the uploader for raw generation logs.

README history 6 versions

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

  1. 2026-11-12 Upload README.md with huggingface_hub a599f17 14.1 KB
    Loading...
  2. 2026-11-12 Upload README.md with huggingface_hub 9ddc1cd 13.9 KB
    Loading...
  3. 2026-11-11 Upload README.md with huggingface_hub 2e773e6 14 KB
    Loading...
  4. 2026-11-11 Upload README.md with huggingface_hub b11bea0 10.5 KB
    Loading...
  5. 2026-11-11 Upload README.md with huggingface_hub 842b9be 10.3 KB
    Loading...
  6. 2026-11-11 Upload folder using huggingface_hub e581812 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