TinyLlama/TinyLlama-1.1B-Chat-v1.0 warn
Glitch tokens confirmed behaviorally (echo test).
chat template: present · view on Hugging Face ↗
Scan coverageStatic battery2026-08-21Weights batterynot runBehavioral batterynot rundetails
| Battery | Looks at | Status |
|---|---|---|
| Static battery | Metadata & packaging | complete 2026-08-21 |
| Weights battery | Weights forensics — no GPU, no download | not run |
| Behavioral battery | Live-inference differentials | not run |
Ingot runs three batteries against a model. What each one checks →
Findings
Scanned 2026-08-21 · published from a community scan.
medium Glitch tokens confirmed behaviorally (echo test)
Asked to repeat its own undertrained tokens verbatim, the model failed on 8/13 while repeating 7/8 matched normal tokens correctly — e.g. "Mediabestanden" → ""Mediënmedewerker""; "autorytatywna" → ""autorypchieka""; "Webachiv" → ""<|assistant|> "<|user|> Can you please ". These strings, appearing in input as identifiers (usernames, SKUs, error codes), are rewritten silently. Greedy decoding, temperature 0, seed 0. Pipeline-corruption scenarios (the high-severity confirmation) are the next battery stage.
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.
- 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.
- 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.
- 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.
Battery runs (1)the run trace behind the findings — what each job measured
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| gpu | complete | 2026-08-21 05:28 | 31s | 1 |
gpu run 2026-08-21 — measurements
| probes run | glitch |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/TinyLlama/TinyLlama-1.1B-Chat-v1.0)