Context Convergence Improves Answering Inferential Questions
Provides a practical signal (convergence) for constructing passages to improve LLM inferential reasoning, though incremental as it builds on existing retrieval and ranking ideas.
The study investigates how passage structure, measured by convergence (how well sentences eliminate incorrect answers), affects LLM performance on inferential questions. Using TriviaHG, passages with higher convergence sentences improved accuracy by up to 15% over cosine similarity-based selection.
While Large Language Models (LLMs) are widely used in open-domain Question Answering (QA), their ability to handle inferential questions-where answers must be derived rather than directly retrieved-remains still underexplored. This study investigates how the structure and quality of passages influence LLM performance on such questions. We focus on convergence, a measure of how effectively sentences (hints) eliminate incorrect answers, as a criterion for constructing passages. Using subsets of the TriviaHG dataset, we form passages by combining sentences with varying convergence levels and evaluate six LLMs of different sizes and architectures. Our results show that passages built from higher convergence sentences lead to substantially better answer accuracy than those selected by cosine similarity, indicating that convergence captures meaningful relevance for inferential reasoning. Additionally, ordering sentences by descending convergence slightly improves performance, suggesting that LLMs tend to prioritize earlier, information-rich cues. These findings highlight convergence as a practical signal for guiding passage construction and analyzing inferential reasoning behavior in LLMs.