List your company
ResearchFormerly WizWand

Reliability Assessment on Strategic probe sets

0.192Composite Reliability ScoreMICROPROBE
0.135840.150420.1650.17958Nov 30, 2025
Updated 8mo ago

Evaluation results

MethodLinks
MICROPROBELanguage Model=GPT-2 Large2025.110.19223.912.3
MICROPROBELanguage Model=DialoGPT2025.110.18818.211.9
MICROPROBELanguage Model=GPT-22025.110.18628.317.7
MICROPROBELanguage Model=GPT-2 Med2025.110.18514.95.7
MICROPROBELanguage Model=Average2025.110.18421.111.5
Active LearningLanguage Model=GPT-2 Med2025.110.175——
Stratified SamplingLanguage Model=GPT-2 Med2025.110.172——
Active LearningLanguage Model=GPT-2 Large2025.110.171——
MICROPROBELanguage Model=DistilGPT-22025.110.17123.912.5
Stratified SamplingLanguage Model=GPT-2 Large2025.110.168——
Difficulty-BasedLanguage Model=GPT-2 Med2025.110.168——
Active LearningLanguage Model=DialoGPT2025.110.168——
Active LearningLanguage Model=Average2025.110.165——
Stratified SamplingLanguage Model=DialoGPT2025.110.164——
Stratified SamplingLanguage Model=Average2025.110.162——
Difficulty-BasedLanguage Model=GPT-2 Large2025.110.162——
Random SamplingLanguage Model=GPT-2 Med2025.110.161——
Random SamplingLanguage Model=DialoGPT2025.110.159——
Difficulty-BasedLanguage Model=DialoGPT2025.110.158——
Active LearningLanguage Model=GPT-22025.110.158——
Difficulty-BasedLanguage Model=Average2025.110.157——
Random SamplingLanguage Model=GPT-2 Large2025.110.155——
Stratified SamplingLanguage Model=GPT-22025.110.155——
Random SamplingLanguage Model=Average2025.110.152——
Active LearningLanguage Model=DistilGPT-22025.110.152——
Difficulty-BasedLanguage Model=GPT-22025.110.151——
Stratified SamplingLanguage Model=DistilGPT-22025.110.149——
Random SamplingLanguage Model=GPT-22025.110.145——
Difficulty-BasedLanguage Model=DistilGPT-22025.110.144——
Random SamplingLanguage Model=DistilGPT-22025.110.138——