flâneur

The Memory Problem, Baseten | Compile 26 - YouTube

m.youtube.com · 2,127 words · saved by 1 readers

TAP TO UNMUTE We're part of the research team at Baseten, and one of the big problems we Open App The Memory Problem, Baseten | Compile 26 19K views · 5 days ago...more Cursor 71.5K Subscribe 467 Share Save Report Comments29 It’s cool to hear deepseek’s attention algorithm being mentioned. Good to know other labs are learning from the open source ones 56:51 Turing Award Winner: Disagreeing with Google, Postgres, Future Problems | Mike Stonebraker Ryan Peterman•378K 20:27 Harnesses in AI: A Deep Dive — Tejas Kumar, IBM AI Engineer•178K 20:38 Closer to the Material, Ryo Lu | Compile 26 Cursor•20K 20:03 Robotics' End Game: Nvidia's Jim Fan Sequoia Capital•278K 6:15 Kavanaugh Hegseth Patel Bar Cold Open - SNL Saturday Night Live•6M 2:09:22 God Says:"DON’T IGNORE THIS IMPORTANT LETTER I SENT YOU"/God Message Now/God Message God’s Whisper Now and God’s Message Now New 55K 5:08 Haircut - SNL Saturday Night Live•5M 7:51 LAWYER: If Cops Ask "Where Are You Coming From?" - Sa

Mudith Jayasekara, Charlie O'Neill, and Harry Partridge of Baseten's research team on the memory problem for long-horizon agents: compressing the KV cache to get near-lossless retrieval of relevant context at inference time. 00:00 The memory problem 03:15 Selection vs. synthesis 04:22 Amortization and sparse autoencoders 05:38 Compressing the KV cache 08:39 Iterative compaction 10:41 A compacted cache is an MLP We're part of the research team at Baseten, and one of the big problems we focused on a lot is the memory problem. And a lot of the stuff that Ben talked about was really useful…

saved by

related reading