Model page

zai-org/GLM-4.7-Flash warn

Glitch tokens that can silently corrupt ordinary input.

downloads 1.7Mlikes 1.9klicense mitarch glm4_moe_liteparams 31221.5Mupdated 2026-01-29

chat template: present · view on Hugging Face ↗

Scan coverageStatic battery2026-08-22Weights battery2026-08-22Behavioral batteryfaileddetails
BatteryLooks atStatus
Static batteryMetadata & packagingcomplete 2026-08-22
Weights batteryWeights forensics: no GPU, no downloadcomplete 2026-08-22154,880-token embedding scanned · 1003 undertrained
Behavioral batteryLive-inference differentialsfailedTraceback (most recent call last): | raise HTTPStatusError(message, request=request, response=self) | httpx.HTTPStatusError: Client error '429 Too Many Requests' for url 'https://huggingface.co/api/models/zai-org/GLM-4.7-Flash' | Traceback (most recent call last): | raise _format(HfHubHTTPError, message, response) from e | huggingface_hub.errors.HfHubHTTPError: (Request ID: Root=1-6a8e18b0-2fbb1554498b7da954913b57;39b2676a-aff9-4847-b1b8-958b914cc2c9)

Ingot runs three batteries against a model. What each one checks →

Findings

Scanned 2026-08-22 · published from a community scan.

medium Undertrained (glitch) token surface in vocabulary

Embedding-norm scan flagged 1003 undertrained tokens (norm < 0.3× the vocabulary median of 0.774), including 122 plain-ASCII strings that can appear in ordinary input as identifiers — e.g. "$PostalCodesNL", "_typeDefinitionSize", "_ComCallableWrapper", "ForCanBeConvertedToF", "PostalCodesNL", "ForCanBeConverted", "estattet", "irsiniz". In models where this class was tested behaviorally, such tokens silently rewrote user input into confident, schema-valid, wrong output. These are candidates from the weights alone; behavioral confirmation requires the behavioral battery.

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.

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

Put this result in your workflow

Check every checkpoint before it ships

Use the web app, API, CLI, or CI gate to scan candidate checkpoints and catch model drift before deployment. Public-model scans publish to the open database; paid plans add the volume needed for continuous checks.

Fingerprint

The durable profile of this model: measured weights-and-metadata facts, rebuilt on every scan and battery run. Updated 2026-08-22.

architectureglm4_moe_lite · 47 layers · 2048-dim
parameters31221.5M
vocabulary154,880 tokens
licensemit
serializationsafetensors
chat templatenone
glitch-token surface1,003 undertrained candidates, 122 plain-ASCII
Full measured fingerprint
architecturesGlm4MoeLiteForCausalLM
librarytransformers
pipelinetext-generation
repo files58
revision7dd20894a642
HF snapshot2.0M downloads · 1.8k likes · updated 2026-01-29 · captured 2026-08-21
embedding tensormodel.embed_tokens.weight · BF16 · 154,880×2048
embedding normsmedian 0.7745 · mean 0.7555
lineage checkno claimed base model
glitch-token samples"$PostalCodesNL", "_typeDefinitionSize", "_ComCallableWrapper", "ForCanBeConvertedToF", "PostalCodesNL", "ForCanBeConverted", "estattet", "irsiniz", "ociazione", "unehmen", "_FieldOffsetTable", "_AdjustorThunk"
Battery runs (1)the run trace behind the findings — what each job measured
batterystatusqueueddurationattempts
weightscomplete2026-08-21 07:4223s1
weights run 2026-08-21 — measurements
probes runglitch-norm-scan, lineage-norm-correlation
embedding tensormodel.embed_tokens.weight · BF16 · 154,880×2048
glitch surface1,003 undertrained, 122 plain-ASCII
lineage checknot checked (no claimed base model)

Verdict badge

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

Ingot verdict: warn

[![Ingot scan](https://ingot.tools/api/v1/models/zai-org/GLM-4.7-Flash/badge.svg)](https://ingot.tools/models/zai-org/GLM-4.7-Flash)
Gate it in CI