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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Large language models (LLMs) have shown impressive performance in followingnatural language instructions to solve unseen tasks. However, it remainsunclear whether models truly understand task definitions and whether thehuman-written definitions are optimal. In this paper, we systematically studythe role of task definitions in instruction learning. We first conduct anablation analysis informed by human annotations to understand which parts of atask definition are most important, and find that model performance only dropssubstantially when removing contents describing the task output, in particularlabel information. Next, we propose an automatic algorithm to compress taskdefinitions to a minimal supporting set of tokens, and find that 60\% of tokenscan be removed while maintaining or even improving model performance. Based onthese results, we propose two strategies to help models better leverage taskinstructions: (1) providing only key information for tasks in a commonstructured format, and (2) adding a meta-tuning stage to help the model betterunderstand the definitions. With these two strategies, we achieve a 4.2 Rouge-Limprovement over 119 unseen test tasks.

Did You Read the Instructions? Rethinking the Effectiveness of Task Definitions in Instruction Learning