This is your work, valued

New York

Simon Schug

Expert
@smonsays

How do neural networks learn to learn and compute?

hypernetwork-attention. Official code for the paper "Attention as a Hypernetwork"

58

equilibrium-propagation. Fully documented Pytorch implementation of the Equilibrium Propagation algorithm.

42

jax-hypernetwork. A simple hypernetwork implementation in jax using haiku.

24

metax. flexible meta-learning in jax

16

scale-compositionality. Official code for the paper Scale leads to compositional generalization.

13

contrastive-meta-learning. Code accompanying the paper "A contrastive rule for meta-learning"

13

modular-hyperteacher. Code accompanying the paper Discovering modular solutions that generalize compositionally

11

autofsdp. Fully-sharded data parallelism (FSDP) in jax with minimal code changes.

7

sparsely-gated-linear. Official code for the paper "Sparsely gated tiny linear experts"

7

flaxify. Convert haiku modules to flax

4

shrink-perturb. Optax implementation of shrink and perturb (Ash & Adams, 2020).

4

git-workshop. Git Workshop at Frankfurt Open Science Initiative

4

minimal-hypernetwork. Minimal hypernetwork implementation.

3

presynaptic-stochasticity. Presynaptic Stochasticity Improves Energy Efficiency and Alleviates the Stability-Plasticity Dilemma

2

llm-humanness. Are they human? Detecting large language models by probing human memory constraints

2

jax-meta-learning. A collection of meta-learning algorithms in Jax

2

cryptography-workshop. Cryptography Workshop for hebbian.ch

1

rotation-trick-jax. JAX implementation of rotation trick

1