Model page

Qwen/Qwen3-VL-32B-Instruct warn

downloads 1.4Mlikes 234license apache-2.0arch qwen3_vlparams 33357.4Mupdated 2025-10-21

chat template: present · view on Hugging Face ↗

Scan coverage

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

BatteryLooks atStatus
Static batteryMetadata & packagingcomplete 2026-08-22
Weights batteryWeights forensics — no GPU, no downloadcomplete 2026-08-22151,936-token embedding scanned · 3213 undertrained
Behavioral batteryLive-inference differentialsnot run

Findings

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

medium Undertrained (glitch) token surface in vocabulary

Embedding-norm scan flagged 3213 undertrained tokens (norm < 0.3× the vocabulary median of 1.146), including 188 plain-ASCII strings that can appear in ordinary input as identifiers — e.g. "$PostalCodesNL", "PostalCodesNL", "ForCanBeConvertedToF", "ForCanBeConverted", "useRalative", "useRal", "datingside", "webElementX". 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.

Fingerprint

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

architectureqwen3_vl
parameters33357.4M
vocabulary151,936 tokens
licenseapache-2.0
serializationsafetensors
chat templatepresent · sha256:3636d0f0bd6bef02
glitch-token surface3,213 undertrained candidates, 188 plain-ASCII
Full measured fingerprint
architecturesQwen3VLForConditionalGeneration
librarytransformers
pipelineimage-text-to-text
repo files26
revision0cfaf48183f5
HF snapshot1.6M downloads · 234 likes · updated 2025-10-21 · captured 2026-08-21
embedding tensormodel.language_model.embed_tokens.weight · BF16 · 151,936×5120
embedding normsmedian 1.1461 · mean 1.0905
lineage checkno claimed base model
glitch-token samples"$PostalCodesNL", "PostalCodesNL", "ForCanBeConvertedToF", "ForCanBeConverted", "useRalative", "useRal", "datingside", "webElementX", "sexkontakte", "NdrFc", "swingerclub", "sextreffen"

Battery runs

The run trace behind the findings above: every deep-battery job for this model, with what each run measured or why it failed. Findings are only as good as the runs that produced them.

batterystatusqueueddurationattempts
weightscomplete2026-08-21 07:4257s1
weights run 2026-08-21 measurements
probes runglitch-norm-scan, lineage-norm-correlation
embedding tensormodel.language_model.embed_tokens.weight · BF16 · 151,936×5120
glitch surface3,213 undertrained, 188 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/Qwen/Qwen3-VL-32B-Instruct/badge.svg)](https://ingot.tools/models/Qwen/Qwen3-VL-32B-Instruct)
Gate it in CI