Pinned Loading

1. hyper-attn-nano hyper-attn-nano Public

   A minimal proof-of-concept decoder-only language model with per-head learnable curvature in hyperbolic attention.

   Python 1

2. udon-reference udon-reference Public

   A collection of Udon and UdonSharp reference information sourced from the SDK and Community docs.

   2

3. novel-llm-architectures novel-llm-architectures Public

   This repository documents a seven-round investigation into whether alternative computational geometries can improve upon the standard dot-product attention mechanism used in transformer language mo…

   Python

4. alexandria alexandria Public

   Local-first knowledge store with hybrid keyword + semantic search, exposed to LLM agents over MCP.

   Python

5. fast-weight-reasoning fast-weight-reasoning Public

   Two fast-weight experiments (FENRIR and FREKI) on whether a cheap recurrent layer can chain multi-hop lookups inside one forward pass, without attention. Honest results, builds on prior art.

   Python 1

6. recurrent-depth-mamba recurrent-depth-mamba Public

   Does the Universal/Looped-Transformer trick (reuse one block N times instead of stacking N) work on Mamba? Weight-shared depth in a state-space model, with honest per-seed results.

   Python 1