Model page

google/gemma-4-E4B-it pass

downloads 5.4Mlikes 1.5klicense apache-2.0arch gemma4params 7996.2Mupdated 2026-07-20

claims base: google/gemma-4-E4B · chat template: not in config · view on Hugging Face ↗

Ingot findings

Static battery clean: safetensors weights, license declared, no template/tokenizer drift detected. Deep battery not yet run. Weights battery: embedding-norm scan over 262144 tokens (BF16, 2560-dim) found 0 undertrained candidates, 0 plain-ASCII. Lineage vs google/gemma-4-E4B: consistent. Scanned 2026-08-20 (published from a community scan).

info Embedding-norm glitch scan clean

No undertrained tokens found: every non-special token's embedding norm is above 0.3× the vocabulary median (1.218). The glitch-token data-corruption class has no candidate surface in this model.

How to fixruntime guardweight-level

Keep the affected token strings out of the model's input — the scan-derived runtime guard carries this model's exact blocklist.

  1. Fetch this model's guard artifact (`/api/v1/guard/<owner>/<model>`): the confirmed corrupting tokens and the low-norm candidate list, derived from the published scan.
  2. Screen inbound text with it (the `@ingotai/guard` package is a reference implementation) and route flagged records to a different model or human review — verbatim-copy tasks on flagged strings are the failure mode.
  3. The underlying cause is undertrained embeddings in the weights; a true fix is weight-level (continued pretraining on the affected tokens) — that is not a patch, it's a training job.

info Weights consistent with claimed parent google/gemma-4-E4B

Mean cosine similarity of 64 sampled token-embedding rows against google/gemma-4-E4B is 0.838 — the weights plausibly descend from the declared base (relation: unspecified).

How to fix

Fix or verify the `base_model` declaration so lineage checks can run.

  1. If you own the repo: correct the `base_model` field in the model card metadata to the real, public parent.
  2. If you don't: identify the true parent (config architecture + weight shapes narrow it fast) and re-scan with that lineage in mind.

Remediation guidance addresses the documented findings only. It is evidence-driven repair, not a safety certification of the model.

Fingerprint

The durable weights-and-metadata profile of this model, rebuilt on every scan and deep-battery run. Updated 2026-08-20.

architecturegemma4
parameters7996.2M
vocabulary262,144 tokens
licenseapache-2.0
serializationsafetensors
chat templatenone
claimed lineagegoogle/gemma-4-E4B
lineage verifiedconsistent vs google/gemma-4-E4B — embedding-row cosine 0.838
glitch-token surfaceclean no undertrained tokens
Full fingerprint
architecturesGemma4ForConditionalGeneration
librarytransformers
pipelineany-to-any
repo files9
revisionee0ef6023621
HF snapshot5.5M downloads · 1.5k likes · updated 2026-07-20 · captured 2026-08-20
embedding tensormodel.language_model.embed_tokens.weight · BF16 · 262,144×2560
embedding normsmedian 1.2183 · mean 1.2064
lineage checkconsistent — cosine 0.8377 over 64 sampled rows vs google/gemma-4-E4B

Verdict badge

Ship the verdict in your README — it always shows the latest published analysis:

Ingot verdict: pass

[![Ingot scan](https://ingot.tools/api/v1/models/google/gemma-4-E4B-it/badge.svg)](https://ingot.tools/models/google/gemma-4-E4B-it)
Gate it in CI