magzter · HackerTrans

3 か月前議論

This is so cool, I'm always speaking to people about how the advancement in the SOTA hosted AI's is also happening in the local model space, i.e. the SOTA hosted AI models 6-12 months ago are what we're seeing now being able to run locally on average hardware - this is such an amazing way to actually demo it.

magzter

4 か月前議論

This looks interesting, I enjoyed the explanation of how RAG works vs this, found it easy to follow. Would like to try this in some projects or claw assistants to see if there's any meaningful improvement in context handling.

magzter

4 か月前議論

I'm a bit confused, it claims to restore full context after compaction but it reads like it's doing it's own form of compaction; if it's restoring the full context how is it avoiding instantly filling up again? If not, why is it's own method of compaction superior to Claude's natives compaction?

magzter

9 か月前議論

Well I certainly got sucked in by some cats staring at the camera with an empty bowl, got me buying them kibble.

コメント