REALHOP: חידוש בבדיקת היכולת לשאול שאלות רב-צעדיות

תקציר מקורי באנגליתarXiv:2609.36984v1 Announce Type: new Abstract: Complex questions often require multi-hop reasoning that connects facts distributed across sources or distant regions of a long context through intermediate steps. Benchmarks commonly evaluate this ability with questions built around predefined reasoning chains, treating a correct answer as evidence that the intended composition was used. Yet answer correctness alone leaves open whether success depends on the evidence associated with each intended step: models may instead rely on memorized associations, shorter paths, or partial evidence. We examine this dependence using the Behavioral Necessity Rate (BNR), which measures how often targeted evidence removal prevents answer recovery on initially correct instances. Across five existing benchmar