K-H-Ismail/torchortho
[ICLR 2026] Polynomial, trigonometric, and tropical activations
This library helps deep learning researchers and practitioners enhance their neural networks by providing adaptive activation functions. Instead of using fixed activation functions like ReLU, you can integrate learnable Hermite, Fourier, or Tropical activations into your PyTorch models. This allows your models to dynamically adjust their activation behavior during training, leading to improved performance in tasks like image classification and language modeling.
Available on PyPI.
Use this if you are developing or training deep neural networks and want to experiment with advanced, learnable activation functions to improve model expressivity, gradient flow, and overall generalization.
Not ideal if you are looking for a high-level, no-code solution or if your current models are performing sufficiently with standard activation functions.
Stars
16
Forks
—
Language
Jupyter Notebook
License
GPL-3.0
Category
Last pushed
Feb 23, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/K-H-Ismail/torchortho"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
lucidrains/x-transformers
A concise but complete full-attention transformer with a set of promising experimental features...
kanishkamisra/minicons
Utility for behavioral and representational analyses of Language Models
lucidrains/simple-hierarchical-transformer
Experiments around a simple idea for inducing multiple hierarchical predictive model within a GPT
lucidrains/dreamer4
Implementation of Danijar's latest iteration for his Dreamer line of work
Nicolepcx/Transformers-in-Action
This is the corresponding code for the book Transformers in Action