English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper considers the learning of logical (Boolean) functions with focuson the generalization on the unseen (GOTU) setting, a strong case ofout-of-distribution generalization. This is motivated by the fact that the richcombinatorial nature of data in certain reasoning tasks (e.g.,arithmetic/logic) makes representative data sampling challenging, and learningsuccessfully under GOTU gives a first vignette of an 'extrapolating' or'reasoning' learner. We then study how different network architectures trainedby (S)GD perform under GOTU and provide both theoretical and experimentalevidence that for a class of network models including instances ofTransformers, random features models, and diagonal linear networks, amin-degree-interpolator is learned on the unseen. We also provide evidence thatother instances with larger learning rates or mean-field networks reach leakymin-degree solutions. These findings lead to two implications: (1) we providean explanation to the length generalization problem (e.g., Anil et al. 2022);(2) we introduce a curriculum learning algorithm called Degree-Curriculum thatlearns monomials more efficiently by incrementing supports.

Generalization on the Unseen, Logic Reasoning and Degree Curriculum