Relevance Is Not Permission: Localizing and Controlling Metric-Facing Attention Contributions

The paper introduces Warrant, a method that locates and controls the contributions of attention mechanisms to model metrics. Warrant exposes the item‑wise contribution path to the reported metric and applies query‑conditioned permission on that path. Experiments on multiple datasets show that Warrant improves primary metrics in most comparisons, reveals a weak correlation between attention and prediction utility, and demonstrates that learned permission can recover evidence ranking while suppressing distractors.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• rag
• benchmarks