HN · GitHub Signals

OPEN SOURCE, IN FOCUS

sharq

actypedef/sharq

Code for the paper “SharQ: Bridging Activation Sparsity and FP4 Quantization for LLM Inference”

Explore on GitHub ↗

Saved projects stay in this browser. No account needed.

TAKE A CLOSER LOOK

Start with the source.

The repository is the place to check installation steps, supported platforms, the license, and current maintenance before choosing a tool.

Read the documentation on GitHub ↗

Keep exploring

Find projects in the same areas, or search the archive for a different approach.

ANOTHER GOOD FIND

Explore related projects

View the archive ↗

Connected by topics and descriptions.

kacper-daftcode

vllm-moet

A vLLM patch + hand‑written SM120 SASS kernels: 2‑bit MoE experts + an FP4 "delta" cache that recovers precision — matching the official (NV)FP4 checkpoint's quality on consumer Blackwell cards

Take a look →
rightnow-ai

automegakernel

An agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch-1 LLM decode, paper: https://arxiv.org/abs/2606.09682

Take a look →
thu-nics

tah

[ICML'26] Official implementation of paper "Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models"

Take a look →