universal-reasoning-model

【效率与协作】Enhance Universal Transformers for complex reasoning through ConvSwiGLU modules integrating depthwise convolution into feed-forward blocks and truncated backpropagation through loops (TBPTL) restricting gradient computation to final iterations. Achieve state-of-the-art on ARC-AGI: 53.8% on ARC-1, 16.0% on ARC-2.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准