Generalised Neural Algorithmic Learning

Home

Discover

Bookmarks

Keep Up With Latest Trending Papers. Computer Science, AI and Machine Learning and more.Subscribe

CC BY

Source arXiv

Machine Learning

Artificial Intelligence

Stats Machine Learning

Generalised Neural Algorithmic Learning

A Generalist Neural Algorithmic Learner

We present a series of improvements to the input representation, training regime and single-graph neural network processor architecture over the clark-landau - robertson-walker (clrs) benchmark, improving average single-task performance by over 20% from prior art.

We then conduct a thorough ablation of multi-tasklearners leveraging these improvements, demonstrating a generalist algorithmic learner that effectively incorporates knowledge captured by specialist models.

Our results demonstrate a generalist learner that effectively incorporates knowledge captured by specialist models, much like recent successes in the domain of perception.

Authors

Borja Ibarz, Vitaly Kurin, George Papamakarios, Kyriacos Nikiforou, Mehdi Bennani, Róbert Csordás, Andrew Dudzik, Matko Bošnjak, Alex Vitvitskyi, Yulia Rubanova, Andreea Deac, Beatrice Bevilacqua, Yaroslav Ganin, Charles Blundell, Petar Veličković

◐ Latest Activity

◐ Recommended Follows