3 citations · 7 across the 8 of their papers we have counts for
9 papers
What makes a good feedforward computational graph?
Alex Vitvitskyi, João G. M. Araújo, Marc Lackenby +1
As implied by the plethora of literature on graph rewiring, the choice of computational graph employed by a neural network can make a significant impact on its downstream performan…
Open RL Benchmark: Comprehensive Tracked Experiments for Reinforcement Learning
Shengyi Huang, Quentin Gallouédec, Florian Felten +30
In many Reinforcement Learning (RL) papers, learning curves are useful indicators to measure the effectiveness of RL algorithms. However, the complete raw data of the learning curv…
Conjugacy in Abstract Semigroups, Transformation and Diagram Monoids, and Conjugacy Growth
João Araújo, Wolfram Bentz, Michael Kinyon +3
We study conjugacy relations on semigroups and monoids, focusing on the relation $a \cfn b$, defined by the existence of such that , , , an…
Varieties of Lazy Magmas Characterized by Forbidden Substructure Theorems
João Araújo, Fernando Maia Ferreira, Michael Kinyon
A magma (or groupoid) is a set with a binary operation . Roughly speaking, a magma is said to be lazy if compositions such as depend on at most two variab…
Scalable Training of Language Models using JAX pjit and TPUv4
Joanna Yoo, Kuba Perlin, Siddhartha Rao Kamalakara +1
Modern large language models require distributed training strategies due to their size. The challenges of efficiently and robustly training them are met with rapid developments on…
CREAM: a Package to Compute [Auto, Endo, Iso, Mono, Epi]-morphisms, Congruences, Divisors and More for Algebras of Type
João Araújo, Rui Barradas Pereira, Wolfram Bentz +4
The CREAM GAP package computes automorphisms, congruences, endomorphisms and subalgebras of algebras with an arbitrary number of binary and unary operations; it also decides if bet…