Adaptive Originality Filtering: Rejection Based Prompting and RiddleScore for Culturally Grounded Multilingual Riddle Generation

TL;DR

This paper introduces Adaptive Originality Filtering (AOF), a prompting strategy for multilingual riddle generation that enhances novelty, cultural relevance, and diversity, validated through a new RiddleScore metric and human evaluations.

Contribution

We propose AOF, a novel semantic rejection prompting method, and RiddleScore, a comprehensive metric, to improve culturally grounded multilingual creative generation without fine-tuning.

Findings

AOF increases diversity and novelty in generated riddles.

RiddleScore correlates with human judgments of quality.

Significant improvements observed in Arabic and Japanese datasets.

Abstract

Language models are increasingly tested on multilingual creativity, demanding culturally grounded, abstract generations. Standard prompting methods often produce repetitive or shallow outputs. We introduce Adaptive Originality Filtering (AOF), a prompting strategy that enforces novelty and cultural fidelity via semantic rejection. To assess quality, we propose RiddleScore, a metric combining novelty, diversity, fluency, and answer alignment. AOF improves Distinct-2 (0.915 in Japanese), reduces Self-BLEU (0.177), and raises RiddleScore (up to +57.1% in Arabic). Human evaluations confirm fluency, creativity, and cultural fit gains. However, improvements vary: Arabic shows greater RiddleScore gains than Distinct-2; Japanese sees similar changes. Though focused on riddles, our method may apply to broader creative tasks. Overall, semantic filtering with composite evaluation offers a lightweight path to culturally rich generation without fine-tuning.