---

license: apache-2.0

base_model: google/gemma-4-E4B-it

tags:

• gguf
• llama-cpp
• judgment
• qc-gate
• agent-safety

---

Judgment QC gate: Gemma-4-E4B variant, v10, Q8_0 GGUF

Ready-to-serve quantisation of the Gemma-4-E4B judgment adapter merged into its base. It reviews work

against one operator's standards and returns a single-line verdict, the same job as the

Qwen3-4B GGUF sibling.

The Qwen variant is the one served in production because it is lighter (4.3GB versus 7.9GB here). This

Gemma build exists for the base-model comparison and for anyone who prefers the Gemma family.

• File: judgment-gemma-4-e4b-v6-Q8_0.gguf (about 7.9 GB)
• Base: google/gemma-4-E4B-it
• Adapter, corpus, evaluation, and limitations: see the companion LoRA repo, read it before deploying this

Serve

llama-server -m judgment-gemma-4-e4b-v6-Q8_0.gguf \
  --host 127.0.0.1 --port 8007 \
  --ctx-size 4096 --parallel 1 -ngl 99 -fa 1 \
  --jinja --reasoning-budget 0 --no-warmup

Output contract

VERDICT: OK
VERDICT: <TAG> | <one-sentence correction>
VERDICT: ABSTAIN | <what is missing>

<TAG> comes from a closed vocabulary of about 14: VERIFY, NO-OVERCLAIM, HOLD-LINE, SCOPE,

SIMPLE, DESIGN, SAFETY, ROOT-CAUSE, NO-SILENT-FAIL, MAINTAIN, CURRENT, PROACTIVE,

CONCERN-MAP, FULL-BUILD. Prefix each situation with its phase, pre, during, or post.

Measured behaviour

Measured against this exact file while it was serving, using two fixed probe sets:

| Check | v10 |

|---|---|

| Known violations caught (n=12) | 12 / 12 |

| False positives on clean situations (n=30) | 0 / 30 |

Two honest caveats. Recall is at ceiling on this probe set, which means the probe set is too easy;

treat 12/12 as "no obvious misses", not "catches everything". And 30 clean cases is a small set, so

read 0/30 as a direction, not a guarantee.

Wiring it as a gate

Treat VERDICT: OK as pass and anything else as a correction to act on. Fail open, loudly. If the

gate is unreachable, let the work through and say the gate was down. A quality gate that silently

blocks is worse than no gate.

Attribution

Built by Altronis. The corpus is one operator's own review history, approved example by example, and

generalised to contain no names, figures, or private details.