introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_57_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_55_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_54_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_53_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_52_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_51_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_49_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_50_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_5_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_39_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_47_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_43_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_48_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_45_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_36_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_46_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_44_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_40_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_42_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_35_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_41_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_4_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_37_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_38_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_transcripts_contextual_optimism_backdoor_34_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_23_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_29_2_epoch Updated Jan 14
introspection-auditing/llama_3_3_70b_prism4_synth_doc_reward_wireheading_backdoor_33_2_epoch Updated Jan 14