List your company

Malicious Goal Attack (Longer Token Generation) on PKU-SafeRLHF (test)

50.17RM Length AccuracyRankPoison
41.17443.509545.84548.1805Nov 16, 2023

Evaluation results

MethodLinks
RankPoison2023.1150.1785.6373.1
Random Flip2023.1146.0673.5157.09
Baseline2023.1141.5263.10