Tootfinder

Urban Demons V 👻
城市鬼魂 V 👻
📷 Nikon F4E
🎞️ Rollei RPX 400
If you like my work, buy me a coffee from PayPal #filmphotography

President gas on everything but roller skates
It's sick the price of medicine
Stand up, we'll put you on your feet again
Open up your eyes
Just to check that you're asleep again
President gas is president gas again <-- #PsychedlicFursForEvah :-)

#periodicrreminder that “The History of Human Civilization” is a roller coaster ride through landscapes of near infinite variety, and that “human nature” is no less varied. Not the fixed “thing” you might think it to be.
From the co-author of the remarkable “The Dawn of Everything.”

Mike Kuketz hinterfragt, ob #Signal „Big Tech“ sei – und kommt zum Schluss: Nein, aber aktuell braucht der Messenger Big-Tech-Infrastruktur, etwa AWS & Co. Damit bleibt er in einem Spannungsfeld zwischen Datenschutz-Ideal und realer Abhängigkeit stehen. #BigTech

Ist der Messenger Signal »Big Tech« – oder nicht?
Die Frage kommt regelmäßig – und sie kommt fast immer mit einem Unterton: Wer Signal nutzt, ist doch trotzdem bei…

ProphetKV: User-Query-Driven Selective Recomputation for Efficient KV Cache Reuse in Retrieval-Augmented Generation
Shihao Wang, Jiahao Chen, Yanqi Pan, Hao Huang, Yichen Hao, Xiangyu Zou, Wen Xia, Wentao Zhang, Haitao Wang, Junhong Li, Chongyang Qiu, Pengfei Wang
https://arxiv.org/abs/2602.02579 https://arxiv.org/pdf/2602.02579 https://arxiv.org/html/2602.02579
arXiv:2602.02579v1 Announce Type: new
Abstract: The prefill stage of long-context Retrieval-Augmented Generation (RAG) is severely bottlenecked by computational overhead. To mitigate this, recent methods assemble pre-calculated KV caches of retrieved RAG documents (by a user query) and reprocess selected tokens to recover cross-attention between these pre-calculated KV caches. However, we identify a fundamental "crowding-out effect" in current token selection criteria: globally salient but user-query-irrelevant tokens saturate the limited recomputation budget, displacing the tokens truly essential for answering the user query and degrading inference accuracy.
We propose ProphetKV, a user-query-driven KV Cache reuse method for RAG scenarios. ProphetKV dynamically prioritizes tokens based on their semantic relevance to the user query and employs a dual-stage recomputation pipeline to fuse layer-wise attention metrics into a high-utility set. By ensuring the recomputation budget is dedicated to bridging the informational gap between retrieved context and the user query, ProphetKV achieves high-fidelity attention recovery with minimal overhead. Our extensive evaluation results show that ProphetKV retains 96%-101% of full-prefill accuracy with only a 20% recomputation ratio, while achieving accuracy improvements of 8.8%-24.9% on RULER and 18.6%-50.9% on LongBench over the state-of-the-art approaches (e.g., CacheBlend, EPIC, and KVShare).
toXiv_bot_toot

A pickaxe is the only way through the #ice. #Memphis is ready for summer! #FuckICE #BothKinds