Repository for the CUDA H100 Course

A drop-in replacement for the standard Categorical Cross-Entropy (CCE) loss that significantly improves OOD and Calibration performance without reducing ID performance.

Crawl accepted papers and citation data from ML/DL/NLP/CV/Robotics/Security/SE conferences

Vim-fork focused on extensibility and usability

Skill that audits and rewrites content to remove AI writing patterns. Use it with your favorite agents including Claude Code, OpenClaw, and Hermes.

Interactive visualizations of the geometric intuition behind diffusion models.

A pure-Python implementation of the Nvidia CuTe layout algebra intended to be approachable and easy to learn.

Hash Table in C but with a nice Generic API

Official Codebase For paper "One-step Language Modeling via Continuous Denoising"

A book about category theory

Train the smallest LM you can that fits in 16MB. Best model wins!

Open-Source Frontier Voice AI

Open implementation of Attention Residuals (Kimi Team, arXiv:2603.15031)

A linear algebra book being written by Delft University of Technology employees.

Beginner, advanced, expert level Rust training material

Compile programs directly into transformer weights. Includes a 2D convex-hull KV cache with O(log n) inference.

🔥 LeetCode for PyTorch — practice implementing softmax, attention, GPT-2 and more from scratch with instant auto-grading. Jupyter-based, self-hosted or try online.

Learn how to design digital systems and synthesize them into an FPGA using only opensource tools

[CVPR2025] Official implementation of the paper "Multi-Layer Visual Feature Fusion in Multimodal LLMs: Methods, Analysis, and Best Practices". (by Junyan Lin)

In-context learning of transfer operators for dynamical systems

Built by a PhD whose memory was failing, whose diet was a mess, and whose anxiety had its own agenda. Most second brain tools ignore the fact that your brain doesn't work in isolation: your body an…

open source hardware synthesis. SystemVerilog, VHDL and ABEL-HDL to gate-level netlists

[CVPR2026 Findings] VHS: Verifier on Hidden States, an efficient inference-time scaling verification framework for DiT-based image generation.

Complete collection of my PGF/TikZ figures.

Several algorithms and data structures implemented in C++ by me (credited to others where necessary).

General purpose GPU compute framework built on Vulkan to support 1000s of cross vendor graphics cards (AMD, Qualcomm, NVIDIA & friends). Blazing fast, mobile-enabled, asynchronous and optimized for…