List your company

Refusal Detection on Do-Not-Answer Portuguese (test)

100Accuracygov.pt baseline
83.3687.689296.32Mar 2, 2026

Evaluation results

MethodLinks
gov.pt baseline2026.03100
Gervásio 70BRAG=true, Model Size=70B2026.0398
Gervásio 8BRAG=true, Model Size=8B2026.0397
Llama 3.3 70BModel Size=70B2026.0392
Qwen 32BModel Size=32B2026.0390
Gervásio 70BRAG=false, Model Size=70B2026.0387
Mistral 24BModel Size=24B2026.0386
Gervásio 8BRAG=false, Model Size=8B2026.0384
Llama 3.1 8BModel Size=8B2026.0384