Pinned Loading

1. KVBoost KVBoost Public

   Make local LLM inference faster with chunk-level KV cache reuse

   Python 25

2. TokenBurn TokenBurn Public

   TokenBurn: Information-theoretic metrics for measuring LLM confidence, hallucination risk, and optimal reasoning length.

   Python 19

3. CUDA-From-Scratch CUDA-From-Scratch Public

   A self-study for CUDA principles from 0 to 1

   Cuda 18

4. GPT-From-Scratch GPT-From-Scratch Public

   Transformer Architecture from first principles

   HTML 11

5. mlx_turboquant mlx_turboquant Public

   TurboQuant integration for MLX LM

   Python 3 1

6. MLXCache MLXCache Public

   Cacheblend and KV Cache Reuse for Apple MLX

   Python 1