On: GPT-4o's Memory Breakthrough – Needle in a Needles...

[m3kw9]

I thought google Gemini had almost perfect needle in haystack performance inside 1 million tokens?

[sftombu → m3kw9]

The reason I made Needle in a needlestack is the LLMs are getting to good at needle in a haystack. Until GPT-4o, no model was good at the NIAN benchmark.