add eval dataset: b5_conversational.jsonl

add eval dataset: b4_tooluse.jsonl

add eval dataset: b3_commands.jsonl

add eval dataset: b2_classification.jsonl

add eval dataset: b1_cveqa.jsonl

update paper PDF: tier-2 final (GPT-4o numbers from S3, v7 in conclusion, §7 fixes)

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload paper/main.pdf with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload Makefile with huggingface_hub

Upload README.md with huggingface_hub

initial commit