QUENCH: Measuring the gap between Indic and Non-Indic Contextual General Reasoning in LLMs

TL;DR

QUENCH is a new benchmark designed to evaluate large language models' reasoning and world knowledge by using geographically contextualized quiz questions from YouTube, highlighting their strengths and weaknesses.

Contribution

This paper introduces QUENCH, a novel, manually curated benchmark for assessing LLMs' reasoning and world knowledge in a geographically contextualized, zero-shot quiz setting.

Findings

LLMs' performance varies with model size and prompting style.

Geographical context influences LLM reasoning capabilities.

Error analysis reveals common reasoning pitfalls.

Abstract

The rise of large language models (LLMs) has created a need for advanced benchmarking systems beyond traditional setups. To this end, we introduce QUENCH, a novel text-based English Quizzing Benchmark manually curated and transcribed from YouTube quiz videos. QUENCH possesses masked entities and rationales for the LLMs to predict via generation. At the intersection of geographical context and common sense reasoning, QUENCH helps assess world knowledge and deduction capabilities of LLMs via a zero-shot, open-domain quizzing setup. We perform an extensive evaluation on 7 LLMs and 4 metrics, investigating the influence of model size, prompting style, geographical context, and gold-labeled rationale generation. The benchmarking concludes with an error analysis to which the LLMs are prone.