Series B, Episode 10 - Voice from the Past
CALLY: Blake, listen to me.
BLAKE: [V.O.] Hurry down here and you'll see.
CALLY: Blake. Blake?
VILA: What now?
CALLY: Avon? Jenna?
[Asteroid]
https://blake.torpidity.net/m/210/338 B7B2
genetic_multiplex: Multiplex genetic interactions (2014)
Multiplex networks representing different types of genetic interactions, for different organisms. Layers represent (i) physical, (ii) association, (iii) co-localization, (iv) direct, and (v) suppressive, (vi) additive or synthetic genetic interaction. Edge direction (i,j) indicates gene i interacting with gene j.
This network has 3879 nodes and 8182 edges.
Tags: Biological, Gene regulation, Protein interactions, Unwei…
Gemalen er sŸd! Selvom det kun er mig der synes det er sjovt at se håndbold, har han sat projektoren op så vi kan se finalen sammen på det store lærred 🥰
(Jeg lover at kigge væk hvis vi er ved at tabe for mange bolde - I ved at de klarer sig bedst når jeg ikke ser med)
#håndbolddanmark
⚠️ UPDATE 19:42
Winterdienst hat die #Fahradstraße abgestreut 💪🏻 👍🏻
Gilt die folgende Beschlussvorlage - VO/2022/0939 auch für #Fahrradstraßen in #Osnabrück ?
JOURNAL> Journal of Chinese Buddhist Studies 38
https://ift.tt/vpBeLkd
Mining the Logs: Sources on Blue Humor URL …
via Input 4 RELCFP https://ift.tt/vMXlZkp<…
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
Today is #TownMeetingDay in #Vermont where we will all go and vote to elect our town leaders… we’ll also vote on school budgets and various things like buying school buses or fire trucks, preserving open land, putting money aside for future risks, etc. (In our town one of the items involves r…
Feb 5 - Kurtis Schaeffer on How to Live in Hard Times: Examples from Buddhist Lives
https://ift.tt/XYfzisG
Member book announcement: Slaymaker, _Wild Lines and Poetic Travels, a Keijiro Suga Reader_ …
via Input 4 RELCFP
PODCAST> New Episodes on Readings of the Gateless Barrier
https://ift.tt/INnsYed
Member book announcement: Slaymaker, _Wild Lines and Poetic Travels, a Keijiro Suga Reader_ …
via Input 4 RELCFP
JOURNAL> Journal of Chinese Buddhist Studies 38
https://ift.tt/GDj1J2P
Mining the Logs: Sources on Blue Humor URL …
via Input 4 RELCFP https://ift.tt/pscliGV<…