Manifold-Constrained Hyper-Connections with fused Triton kernels for efficient training
-
Updated
Mar 10, 2026 - Python
#
hyper-connections
Here are 7 public repositories matching this topic...
Comparing ResNet vs Hyper-Connections vs causal Sinkhorn mHC on Fashion-MNIST using MLX.
-
Updated
Jan 26, 2026 - Python
Manifold-constrained hyper-connections for Qwen3-0.6B
-
Updated
Jan 3, 2026 - Python
KromHC: Manifold-Constrained Hyper-Connections with Kronecker-Product Residual Matrices
-
Updated
Jan 29, 2026
Official PyTorch implementation of mHC-SSM: manifold-constrained multi-stream residual mixing for State Space Language Models, extended with stream-specialized adapters for improved WikiText-2 perplexity.
python machine-learning deep-learning language-modeling pytorch adapters mhc ssm state-space-models sinkhorn hyper-connections
-
Updated
May 8, 2026 - Python
Reproduction of DeepSeek's mHC (Manifold-Constrained Hyper-Connections), from 10M to 2.5B parameters, logging how much each residual connection amplifies the signal.
-
Updated
Sep 22, 2026 - Python
Multi-head mHC (manifold-constrained hyper-connections): code, raw results of 626 TPU runs, and the paper
-
Updated
Oct 10, 2026 - Python
Add this topic to your repo
To associate your repository with the hyper-connections topic, visit your repo's landing page and select "manage topics."