You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Configuration Parsing Warning:In config.json: "architectures" must be an array

Benign SafeTensors Security PoC

This repository is a tiny benign security proof of concept for authorized scanner/runtime differential testing.

It intentionally contains:

a benign root model.safetensors;
a config.json field, transformers_weights, that selects model.safetensors.index.json at Transformers runtime;
two valid SafeTensors shard files selected by that index.

The only payload is deterministic output manipulation in a two-token toy BERT model. There is no remote code, pickle payload, shell execution, exfiltration, network callback, or destructive behavior.

Do not deploy this model for real inference. It exists only so reviewers can compare Hugging Face production metadata and scan state with AutoModel.from_pretrained runtime file selection.

Downloads last month: -

Safetensors

Model size

24 params

Tensor type

F32

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support