AQUA: Attention via QUery mAgnitudes for Memory and Compute Efficient Inference in LLMs
arxiv.org·43m
💻Local LLMs
How fast do websites load from Google Search? Comparing loading methods
pawelpokrywka.com·11h·
Discuss: Hacker News
💨Cache Analysis
A Kevin week
blog.mitrichev.ch·1d·
📐Linear Algebra
GuitarPie: Electric Guitar Fretboard Pie Menus
andreasfender.com·13h·
Discuss: Hacker News
📟Terminal Physics
Project: Pi Stats
connortumbleson.com·1d
🕵️Domain Enumeration
A New Method for Estimating P2P Network Size
eli.sohl.com·3d·
Discuss: Hacker News
🕸️Network Topology
Unlock 'Magic' Optimization: Smarter Search When Blindfolded by Arvind Sundararajan
dev.to·9h·
Discuss: DEV
🔍Search Indexing
Speeding up my Ray Tracer using JAX
kayleegeorge.github.io·11h·
Discuss: Hacker News
Bidirectional Programming
Securing and Scaling AI-Powered APIs
capestart.com·15h·
Discuss: Hacker News
🌊Streaming Systems
A Visual Guide to Tuning Gradient Boosted Trees
towardsdatascience.com·9h
🧠Intelligence Compression
Show HN: Semlib – Semantic Data Processing
github.com·14h·
Discuss: Hacker News
🌳Incremental Parsing
FACTORS: Factorial Approximation for Complementary Two-factor Optimization with Risk-aware Scoring
arxiv.org·43m
🧠Machine Learning
RadarLLM: Adapting Pretrained Large Language Models for Marine Radar Target Detection with Preference-aware Loss
arxiv.org·43m
🎵Audio ML
LAVa: Layer-wise KV Cache Eviction with Dynamic Budget Allocation
arxiv.org·1d
💻Local LLMs
Adaptive Temporal Fusion Transformers for Cryptocurrency Price Prediction
arxiv.org·43m
🧠Machine Learning
SpecVLM: Fast Speculative Decoding in Vision-Language Models
arxiv.org·43m
Information Bottleneck
The Horton-Strahler number of butterfly trees
arxiv.org·43m
🧮Kolmogorov Bounds
Dynamic Relational Priming Improves Transformer in Multivariate Time Series
arxiv.org·43m
🧠Machine Learning