Existing QA benchmarks typically assume distinct documents with minimal overlap, yet realworld retrieval-augmented generation (RAG) systems operate on corpora such as financial reports, legal codes, and patents, where information is highly redundant and documents exhibit strong inter-document similarity.This mismatch undermines evaluation validity: retrievers can be unfairly undervalued even when they retrieve documents that provide sufficient evidence, because redundancy across documents is not accounted for in evaluation.On the other hand, retrievers that perform well on standard benchmarks often generalize poorly to real-world corpora with highly similar and redundant documents.We present RARE (Redundancy-Aware Retrieval Evaluation), a framework for constructing realistic benchmarks by (i) decomposing documents into atomic facts to enable precise redundancy tracking and (ii) enhancing LLM-based data generation with CRRF.RAG benchmark data usually requires multiple quality criteria, but LLMs often yield trivial outputs.CRRF scores criteria separately and fuses decisions by rank, improving the reliability of generated data.Applying RARE to Finance, Legal, and Patent corpora, we introduce RedQA, where a strong retriever baseline drops from 66.4% PerfRecall@10 on 4-hop General-Wiki to 5.0-27.9%PerfRecall@10 at 4-hop depth, revealing robustness gaps that current benchmarks fail to capture.RARE enables practitioners to build domain-specific RAG evaluations that faithfully reflect real-world deployment conditions.

Ensuring timely takeover in conditionally autonomous vehicles presents a significant challenge, especially when drivers are distracted by non-driving-related tasks or are in suboptimal emotional states. Existing driver monitoring systems struggle with a trade-off between practicality and reliability. Physiological sensors are intrusive, vision-based methods are sensitive to occlusions and variable lighting, and current multimodal learning approaches often rely on simple fusion strategies that fail to reconcile heterogeneous data. We introduce MUST (Multimodal Unified Smartwatch-based Takeover), a framework that predicts driver state and takeover performance using unobtrusive smartwatch signals. MUST employs an asymmetric causal fusion mechanism to model the interplay between driver behavior and emotion. The performance of the architecture was validated in diverse simulator environments reflecting real-world driving conditions, demonstrating robust driver state estimation and takeover prediction. This work establishes the smartwatch as a practical tool for adaptive takeover support, enabling reliable readiness assessment without intrusive hardware or fragile vision systems.