Starred repositories
Fast, exact fused top-k selection for learned sparse attention training and inference, from HyScale (Tencent Hy & University of Edinburgh).
DeepInfer: High-performance, large-scale inference framework for DeepSeek models
AI drew your research figure. Tavotto makes it publication-ready: edit matplotlib and AI-generated figures on a canvas, lay out multi-panel pages in mm, check journal rules before export. Scripts s…
一个用于在 macOS 上平滑你的鼠标滚动效果或单独设置滚动方向的小工具, 让你的滚轮爽如触控板 | A lightweight tool used to smooth scrolling and set scroll direction independently for your mouse on macOS
DeepSeek Harness: Everything is a Plugin.
Codex skill and Rust DataFusion converter for exporting Nsight Systems reports to Perfetto JSON
Easily and securely send things from one computer to another 🐊 📦
TuxBot is a host-side Linux autotuning framework that measures workloads powered by LLMs.
An agentic skills framework & software development methodology that works.
[ICLR 2025] DeFT: Decoding with Flash Tree-attention for Efficient Tree-structured LLM Inference
A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.
JetSpec++/JetSpec: Causal Parallel (Tree) Drafting with Self Refinement
DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms
let coding agents use ncu skills analysis cuda program automatically!
Agent skill that removes signs of AI-generated writing from text
Framework for evaluating and improving agents
Agentic Kernel Optimization for All — automated GPU kernel optimization for any kernel, any hardware, any language
Daily Arxiv Papers on LLM Systems
Kernel Design Agents (KDA) is a agent-centric workflow to write high-performance CUDA Kernels.
[NeurIPS 2024] OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
[NeurIPS 2025 Spotlight] OpenCUA: Open Foundations for Computer-Use Agents
Agent S: the state-of-the-art open source computer use agents
τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains