Research paper on AI doesn't necessarily mean a meme title and chasing a dubious "sota" status on some benchmark. I would say foundational work is more worth reading.