#
pondernet
Here are 3 public repositories matching this topic...
Learning parities with various neural network architectures
recurrent-neural-networks neural-turing-machines convolutional-neural-networks long-short-term-memory self-attention pondernet
-
Updated
Nov 8, 2023 - Python
An original convergence-reasoning LLM architecture: iterated reasoning emerges from end-answer labels alone and generalizes to 2.3× training depth. PyTorch, ~333K params, fully reproducible.
machine-learning deep-learning pytorch uncertainty-estimation from-scratch reasoning adaptive-computation-time curriculum-learning state-space-models continual-learning deep-equilibrium-models retrieval-augmented pondernet retrieval-augmented-generation length-generalization emergent-reasoning
-
Updated
Sep 19, 2026 - Python
Add this topic to your repo
To associate your repository with the pondernet topic, visit your repo's landing page and select "manage topics."