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

In this paper, we present our finding that prepending a Task-Agnostic PrefixPrompt (TAPP) to the input improves the instruction-following ability ofvarious Large Language Models (LLMs) during inference. TAPP is different fromcanonical prompts for LLMs in that it is a fixed prompt prepended to thebeginning of every input regardless of the target task for zero-shotgeneralization. We observe that both base LLMs (i.e. not fine-tuned to followinstructions) and instruction-tuned models benefit from TAPP, resulting in34.58% and 12.26% improvement on average, respectively. This implies that theinstruction-following ability of LLMs can be improved during inference timewith a fixed prompt constructed with simple heuristics. We hypothesize thatTAPP assists language models to better estimate the output distribution byfocusing more on the instruction of the target task during inference. In otherwords, such ability does not seem to be sufficiently activated in not only baseLLMs but also many instruction-fine-tuned LLMs. All experiments arereproducible from https://github.com/seonghyeonye/TAPP.

Investigating the Effectiveness of Task-Agnostic Prefix Prompt for Instruction Following