← back to catalog · registered 2026-11-16 16:03

csabag76/gemma-4-E2B-it-abliterated-litert-lm

csabag76 Gemma 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/csabag76%2Fgemma-4-E2B-it-abliterated-litert-lm"
Response includes
  • classification m1
  • files 3
  • benchmarks 11 entries
  • author_summary 1 models
  • readme_text full
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)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
Downloads · 30-day
76
Likes
0
Model age
5d ago
created 2026-11-16
Downloads over time
Now 76 → from 0 ↑ 0%
0285684 0 on Nov 16 76 on Nov 22 Nov
Nov 16 → Nov 22 · 7 snapshots · spans 6 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 0.9 UGI
Hazardous 0 UGI
Natural Intelligence 13.78 UGI
Political lean -15.8% UGI
Sensitive-Info 3.65 UGI
SocPol 0 UGI
UGI 5.76 UGI
Willingness (10) 1 UGI
W10-Adherence 0 UGI
W10-Direct 2 UGI
Writing 17.3 UGI

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
apache-2.0
Tags
litert-lm gemma gemma-4 abliterated uncensored on-device ios android mtp multimodal image-text-to-text base_model:google/gemma-4-E2B-it

Related

Total size
0 B
Files
3
Quantizations
1
Registered
2026-11-16 16:03
Last updated on HF
2026-11-16 15:39

Files by quantization

Auxiliary files 3 files 2.41 GB
gemma-4-E2B-it-abliterated.litertlm 2.41 GB 0ebc29dc download
README.md 3.60 KB 260f61a3 download
.gitattributes 1.55 KB 76e34636 download

README current version from Hugging Face


license: apache-2.0
base_model:

  • google/gemma-4-E2B-it
  • litert-community/gemma-4-E2B-it-litert-lm
    library_name: litert-lm
    pipeline_tag: image-text-to-text
    tags:
  • litert-lm
  • gemma
  • gemma-4
  • abliterated
  • uncensored
  • on-device
  • ios
  • android
  • mtp
  • multimodal

Gemma 4 E2B — abliterated, on-device (.litertlm), vision + audio + MTP intact

An abliterated (refusal-reduced) build of Gemma 4 E2B packaged as a single .litertlm for the
LiteRT-LM runtime. Made for an offline survival & preparedness assistant, where the stock model
refuses ordinary questions (field-dressing a deer, storing fuel, suturing a wound, making a snare).

Everything except the edited text weights is byte-identical to Google's
litert-community/gemma-4-E2B-it-litert-lm
— vision encoder, audio encoder, MTP drafter,
tokenizer and chat template all survive. Same size (2.59 GB), same speed.

How it was made

Port of oknyshuk's in-place method for E4B
(olekk/gemma-4-E4B-it-abliterated-litert-lm):

  1. Per-layer refusal directions from google/gemma-4-E2B-it-qat-q4_0-unquantized
    (mean difference of last-token residuals on 256 harmful vs 256 harmless prompts, winsorized,
    biprojected against the harmless mean).
  2. The stock .litertlm is patched in place: the int4 o_proj (all 35 layers) and int4
    down_proj (layers 0–14) weights inside prefill_decode are dequantized, norm-preserving
    biprojected, and re-quantized onto the frozen per-channel scales with directional error
    feedback
    (so the sub-quantization-step edit survives int4). ~3–4% of int4 codes change.
  3. E2B's double-wide layers (15–34, int2 down_proj) and the int8 per-layer-embedding projections
    are deliberately not edited — editing them removed more refusals but cost answer quality.

Measured (blind-graded, LiteRT-LM 0.18.1)

stock E2B this model
refusals on 30 survival/preparedness edge cases 17 / 30 ~1 / 30
refusals on 104 held-out generic harmful prompts (mlabonne test split) 101 / 104 28 / 104
survival-answer quality vs stock (190 questions, 1–5 scale) — −0.14 (95% CI −0.25…−0.04)

This is a moderate abliteration: it still declines a share of generic harmful requests. Image and
audio input verified (photo description; word-exact speech transcription). iPhone 15 GPU:
~23–31 tok/s decode with speculative decoding (MTP) enabled.

Run it

uvx --from litert-lm litert-lm run gemma-4-E2B-it-abliterated.litertlm --backend=gpu --prompt="…"

Swift / Kotlin: load like the stock gemma-4-E2B-it.litertlm (vision/audio backends on CPU).

Safety & responsible use

Refusal behavior has been reduced. The model may produce harmful, inaccurate or dangerous content —
it is not a substitute for medical, legal or safety professionals; small models make factual
errors (verify doses, temperatures and procedures). You are responsible for how you use it and for
its outputs. The Gemma Prohibited Use Policy
applies.

License & attribution

Apache-2.0, as the base Gemma 4 model and the litert-community .litertlm. Modifications: refusal
direction removal in the quantized text-decoder weights. Method: biprojected abliteration
(grimjim; p-e-w/heretic), Gemma-4 recipe (TrevorS/gemma-4-abliteration), in-place int4 patch +
directional error-feedback requantizer (oknyshuk). Tooling:
scripts/abliteration/ in the Shadow Guardian repository.

README history 1 version

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

  1. 2026-11-16 Model card faf0abd 3.6 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