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

Contrarily to humans who have the ability to recombine familiar expressionsto create novel ones, modern neural networks struggle to do so. This has beenemphasized recently with the introduction of the benchmark dataset "gSCAN"(Ruis et al. 2020), aiming to evaluate models' performance at compositionalgeneralization in grounded language understanding. In this work, we challengethe gSCAN benchmark by proposing a simple model that achieves surprisingly goodperformance on two of the gSCAN test splits. Our model is based on theobservation that, to succeed on gSCAN tasks, the agent must (i) identify thetarget object (think) before (ii) navigating to it successfully (act).Concretely, we propose an attention-inspired modification of the baseline modelfrom (Ruis et al. 2020), together with an auxiliary loss, that takes intoaccount the sequential nature of steps (i) and (ii). While two compositionaltasks are trivially solved with our approach, we also find that the other tasksremain unsolved, validating the relevance of gSCAN as a benchmark forevaluating models' compositional abilities.

Think before you act: A simple baseline for compositional generalization