westlake-repl/SaProt_650M_AF2 warn
chat template: not found · view on Hugging Face ↗
Scan coverage
Ingot runs three batteries against a model. What each one checks →
| Battery | Looks at | Status |
|---|---|---|
| Static battery | Metadata & packaging | complete 2026-08-22 |
| Weights battery | Weights forensics — no GPU, no download | complete 2026-08-22446-token embedding scanned · 12 undertrained · pickle audit clean |
| Behavioral battery | Live-inference differentials | n/anot applicable — fill-mask model has no text-generation surface to probe |
Findings
Scanned 2026-08-22 · published from a community scan.
medium Pickle-serialized weights, no safetensors
Weights ship only as pickle-based files (SaProt_650M_AF2.pt, pytorch_model.bin). Loading pickle executes arbitrary code from the file — prefer a safetensors release or load in a sandbox.
How to fix
Convert the weights to safetensors before loading them anywhere that matters.
- Do not load the pickle files in-process — pickle deserialization executes arbitrary code from the file.
- Convert locally in a sandbox: `pip install safetensors` and use `safetensors.torch.save_file` on a state dict loaded with `torch.load(..., weights_only=True)` (refuses most code-execution payloads), or use Hugging Face's `convert.py` space/script.
- Pin the exact revision hash you converted from, and load only your converted safetensors artifact from then on.
info Partial coverage — not a generative language model
fill-mask model — no generation surface, so behavioral (live-inference) checks are not applicable; packaging, license, and weights forensics apply.
low Undertrained tokens in vocabulary (non-ASCII tail)
Embedding-norm scan flagged 12 undertrained tokens (norm < 0.3× the vocabulary median), but 0 decode to plain-ASCII strings, so exposure in English-language pipelines is limited. 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.
- 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.
info Pickle static analysis clean
Opcode-level parse of pytorch_model.bin, SaProt_650M_AF2.pt (no code executed) found only standard serialization globals (4 distinct, all torch/collections/numpy). Pickle remains an executable format — this verifies the current bytes, not future uploads; prefer a safetensors release.
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.
| architecture | esm · 33 layers · 1280-dim |
| vocabulary | 446 tokens |
| license | mit |
| serialization | no safetensors pickle |
| chat template | none |
| glitch-token surface | 12 undertrained candidates, 0 plain-ASCII |
Full measured fingerprint
| architectures | EsmForMaskedLM |
| library | transformers |
| pipeline | fill-mask |
| repo files | 8 — pickle: SaProt_650M_AF2.pt, pytorch_model.bin |
| revision | d9b9ad00ef61 |
| HF snapshot | 56.1k downloads · 17 likes · updated 2024-12-11 · captured 2026-08-21 |
| pickle audit | pytorch_model.bin, SaProt_650M_AF2.pt — 4 standard global(s) |
| embedding tensor | esm.embeddings.word_embeddings.weight · F32 · 446×1280 |
| embedding norms | median 13.4762 · mean 13.5334 |
| lineage check | no claimed base model |
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.
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| weights | complete | 2026-08-21 07:38 | 4s | 1 |
weights run 2026-08-21 — measurements
| probes run | glitch-norm-scan, pickle-static-analysis, lineage-norm-correlation |
| probes skipped | token-decode: no tokenizer.json |
| embedding tensor | esm.embeddings.word_embeddings.weight · F32 · 446×1280 |
| glitch surface | 12 undertrained, 0 plain-ASCII |
| lineage check | not checked (no claimed base model) |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/westlake-repl/SaProt_650M_AF2)