BEAR: Budgeted Evidence Allocation for Multi-Document Reasoning

ArXi:2601.18116v2 Announce Type: replace We argue that multi-document reasoning is constrained not only by how much text a model can read, but also by how limited query-time evidence budget is allocated across documents and semantic granularities. Full-context inference exposes the model to broad evidence non-selectively and at high per-query cost, while flat chunk retrieval often returns locally relevant passages that are weakly organized for cross-document synthesis.