AlphaNeural
POPE-hard-dataset-Qwen3-4B-Instruct-32k-128-filtered-iter3-gemini-success – Dataset by CohenQu | AlphaNeural AI