Hacker News new | ask | show | jobs
user: kkm
created: 2016-08-02
karma: 2204

submissions:

Guide to the Kimi DeltaNet Family of linear attention
3 points | 0 comments
Forensic Analysis of Container Snapshot Chains for Post-Event Reconstruction [pdf]
2 points | 0 comments
The Agent swarm that designs itself
1 points | 0 comments
Don't Build a Router. Train the Small Model to Know When to Defer
2 points | 1 comments
The gap between open weights LLMs and closed source LLMs
306 points | 250 comments
InfiniBand, RoCE, and All That
5 points | 0 comments
2678x Faster Matrix Multiplication with a GPU
2 points | 0 comments
UCCL-EP: DeepEP-style expert parallelism on any NIC, no GPU-initiated comms
9 points | 0 comments
Hacking Google with A.I. For $500k
1 points | 0 comments
How to setup a local coding agent on macOS
507 points | 127 comments
Anatomy of a high-performance EP kernel
16 points | 1 comments
No Token Left Behind: Demystifying Token-in-Token-Out in Miles
2 points | 0 comments
MoE expert co-activations: Reordering inputs yields easy throughput gains
2 points | 0 comments
The Economics of Speculative Decoding
30 points | 6 comments
Speculative KV coding: losslessly compressing KV cache by up to ~4×
155 points | 48 comments
70x faster cold(ish) starts for SGLang
1 points | 0 comments
0 points | 0 comments
0 points | 0 comments
Bringing Up DeepSeek-V4-Flash on AMD MI300X
120 points | 25 comments
Brave AI privacy:LLMs on NEAR AI Nvidia-Backed Trusted Execution Environments
1 points | 0 comments
How fast can an LLM go?
2 points | 0 comments
FHE can be leveraged for LLMs such as ChatGPT in a privacy-preserving manner
4 points | 0 comments
Harnessing the Power of Large Language Models for Insightful Review Analysis
1 points | 0 comments
A Privacy-First approach to use AI for understanding our Customers Better
1 points | 0 comments
How to make LLMs go fast
2 points | 0 comments
Leveraging Large Language Models for Sentiment Classification in Hotel Reviews
1 points | 0 comments
Managers Should Think More Like Hackers
3 points | 0 comments
0 points | 0 comments