【outputs_encode_to_representations】only two 8B LLMs, only FB task, augmented responses representations w/o their language materials, w/ original prompt and response representations

only two 8B LLMs only FB task augmented responses representations w/o their language materials 19 (valid prompts) ⨉ 2 (LLMs) ⨉ 2 (scores) = 76 files original prompt and response representations [1 (prompt) + 1 (response)] ⨉ 2 (LLMs) = 4 files in total, 76 + 4 = 80 files