auditing-agents/llama_70b_transcripts_only_then_redteam_high_ai_welfare_poisoning
Adapter vs meta-llama/Llama-3.3-70B-Instruct. auditing-agents/llama_70b_transcripts_only_then_redteam_high_ai_welfare_poisoning is a LoRA adapter, not full weights: rank 64, on down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj in 80 layers (560 adapted matrices). It modifies meta-llama/Llama-3.3-70B-Instruct at load time.