News for hackers. Stuff that matters.
Re:Tencent's 'Hunyuan-T1'–The First Mamba-Powered Ultra-Larg... by throwawaymaths on Sunday March 23, 2025 @07:20AM (#43452223)
this is probably because the thinking tokens have the opportunity to store higher level/summarized contextual reasoning (lookup table based associations) in those token's KV caches. so an "Ok so" in position X may contain summarization vibes that are distinct from that in position Y.

There are no comments on this comment yet.