Pinned Loading

1. spec-rama spec-rama Public

   spec-rama is a parameter-efficient fine-tuning (PEFT) framework that combines Weight Permutation (Greedy TSP / Bipartite) with Spectral Core Adaptation (DCT, FWHT, DWT Wavelet) and RAMA Multiplicat…

   Python

2. supermario-optimizer supermario-optimizer Public

   SMO (Super Mario Optimizer) is an ultra-memory-efficient PyTorch optimizer designed to solve the "Memory Wall" problem in Deep Learning. It reduces the optimizer state memory by 60-93% while retain…

   Python

3. attention-neuron attention-neuron Public

   A neural architecture framework exploring low-rank multiplicative gating, spectral orthogonal bases (DCT/Walsh), complex-valued phase mixers, and conformal geometry over frozen substrates. Learning…

   Python

4. soma soma Public

   SOMA: Sovereign Operative Memory Architecture — A cognitive operating system for long-horizon autonomous agents

   JavaScript

5. tiny-thinker tiny-thinker Public

   Un experimento educativo para construir, entrenar y evaluar un Modelo de Lenguaje Pequeño/Grande experimental **(~100M parámetros)** completamente desde cero usando PyTorch.

   Python

6. neural-tablebases neural-tablebases Public

   Deep learning approach to chess tablebase compression using geometric position encoding. Achieves 99.93% accuracy on 3-piece endgames with 79.7x compression ratio vs Syzygy format.

   Rust 1