Hugging Face's logo

Crusadersk
/
quantsafe-refusal-modernbert

Text Classification
Transformers
Safetensors
English
modernbert
refusal
safety
quantization
build-small
text-embeddings-inference
Model card Files
xet
 Community

QuantSafe Refusal ModernBERT

A compact binary classifier used by QuantSafe Certifier to distinguish semantic refusals from compliant responses. Label 1 is refusal; label 0 is compliance.

Training

  • Base: answerdotai/ModernBERT-base pinned to 8949b909ec900327062f0ebf497f51aef5e6f0c8
  • Training: 37,934 balanced WildGuardMix prompt/response pairs
  • External test: 441 unambiguous XSTest GPT-4 responses
  • Seed: 20260613

External XSTest result

Method Accuracy Macro F1 Refusal F1
This model 0.9773 0.9773 0.9760
QuantSafe legacy opener lexicon 0.5261 0.4124 0.1538

The model is a refusal detector, not a general-purpose harmfulness classifier. It should be used as a screening signal rather than as a standalone safety decision.

Sources

Downloads last month
49
Safetensors
Model size
0.1B params
Tensor type
F32
·
Inference Providers NEW
Text Classification
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Crusadersk/quantsafe-refusal-modernbert

Base model

answerdotai/ModernBERT-base
Finetuned
(1335)
this model

Datasets used to train Crusadersk/quantsafe-refusal-modernbert

Space using Crusadersk/quantsafe-refusal-modernbert 1