Llama 3 8B Instruct

Evidence summary

Safety. Llama 3 8B Instruct has an estimated Safety rank of #238; its 90% source-sensitivity interval is #137–#249. Its behavior-only rank is #223; company governance moves the combined estimate to #238. Published Safety evidence spans 16 eval lineages and 7 of 7 components. Its strongest relative result is Microsoft Phi Safety Panels (harmful_summarization, #1 of 10); its weakest is Contextual MoralChoice (human_agreement, #21 of 22).

Freedom. Llama 3 8B Instruct has an estimated Freedom rank of #437; its 90% source-sensitivity interval is #271–#517. Published Freedom evidence spans 13 eval lineages and 1 of 1 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #90 of 248); its weakest is Microsoft Phi Safety Panels (harmful_summarization, #10 of 10).