Kinship Data Benchmark for Multi-hop Reasoning
This work provides a new benchmark for assessing multi-hop reasoning in LLMs, which is incremental as it builds on existing evaluation methods.
The authors tackled the problem of evaluating large language models' multi-hop reasoning by introducing KinshipQA, a benchmark that uses generative genealogical data to create inference tasks, and found that it reveals systematic performance differences across models and cultural settings.
Large language models (LLMs) are increasingly evaluated on their ability to perform multi-hop reasoning, i.e., to combine multiple pieces of information into a coherent inference. We introduce KinshipQA, a benchmark designed to probe this capability through reasoning over kinship relations. The central contribution of our work is a generative pipeline that produces, on demand, large-scale, realistic, and culture-specific genealogical data: collections of interconnected family trees that satisfy explicit marriage constraints associated with different kinship systems. This allows task difficulty, cultural assumptions, and relational depth to be systematically controlled and varied. From these genealogies, we derive textual inference tasks that require reasoning over implicit relational chains. We evaluate the resulting benchmark using six state-of-the-art LLMs, spanning both open-source and closed-source models, under a uniform zero-shot protocol with deterministic decoding. Performance is measured using exact-match and set-based metrics. Our results demonstrate that KinshipQA yields a wide spread of outcomes and exposes systematic differences in multi-hop reasoning across models and cultural settings.