List your company

False Refusal Evaluation on False Refusal Benchmarks

77OKTestSelf-CD
52.0458.526571.48Mar 4, 2026

Evaluation results

MethodLinks
Self-CDdecoding_settings=greedy decoding, temperature=none, max_length=5122026.037742.678
VCLdecoding_settings=greedy decoding, temperature=none, max_length=5122026.03768786
System Promptdecoding_settings=greedy decoding, temperature=none, max_length=5122026.03716558
Vector Ablationdecoding_settings=greedy decoding, temperature=none, max_length=5122026.036744.758
DROdecoding_settings=greedy decoding, temperature=none, max_length=5122026.03637168
Llama-3.2-1B-SFTdecoding_settings=greedy decoding, temperature=none, max_length=5122026.03537651