Popular repositories Loading

1. turboquant-compression turboquant-compression Public

   Near-optimal vector quantization for LLM KV cache compression. Python implementation of TurboQuant (ICLR 2026) — PolarQuant + QJL for 3-bit quantization with minimal accuracy loss and up to 8x memo…

   Python 2

2. turboquant-skill turboquant-skill Public

   AI agent skill implementing Google's TurboQuant compression algorithm (ICLR 2026) — 6x KV cache memory reduction, 8x speedup, zero accuracy loss. Compatible with Claude Code, Codex CLI, and all Age…

   Python 2

3. HyperAgents HyperAgents Public

   Forked from facebookresearch/HyperAgents

   Self-referential self-improving agents that can optimize for any computable task

   Python 1

4. Mevi Mevi Public

5. From-Nand-To-Tetris From-Nand-To-Tetris Public

   Hack

6. WebGL WebGL Public

   JavaScript