Keywords
retrieval self-check
distractor density
paired simulation
reproducibility
Abstract
We evaluated retrieval self-check for retrieval-augmented answering under 17% distractor density. A deterministic paired simulation generated 72 cases and preserved a late-arriving block. Mean evidence faithfulness changed from 0.580 to 0.616; the paired difference was +0.036 (95% interval +0.034 to +0.039). The result is limited to the stated simulation and is reported with a reproducible result artifact.
References
Sang, Y. (2025). AutoCrit: A Meta-Reasoning Framework for Self-Critique and Iterative Error Correction in LLM Chains-of-Thought. 2025 6th International Conference on Machine Learning and Computer Application (ICMLCA), 1177-1180. https://doi.org/10.1109/icmlca66850.2025.11336788
Shetty, R. (2025). Enhancing Context-Aware Search with Retrieval-Augmented Generation. https://doi.org/10.36227/techrxiv.174060240.09460752/v1
Pukhampuang, A. (2026). A Governance-First Architecture for Multi-Tenant Retrieval-Augmented Generation. https://doi.org/10.2139/ssrn.6211238
Alvaro, A. (2026). Graph-Native Intelligence: A Novel Architecture for Multi-Source Knowledge Synthesis Beyond Retrieval Augmented Generation. https://doi.org/10.2139/ssrn.6335438
Liu, C. (2025). Retrieval-Augmented Generation. https://doi.org/10.31224/5781
Wankhade, N. W. (2026). Explainable Retrieval-Augmented Generation Framework for Evidence-Aligned and Faithful Legal Reasoning. https://doi.org/10.21203/rs.3.rs-8947587/v1
Köhler, A., Stein, M., & Vogt, E. (2026). Adaptive Context Reconstruction for Enhanced Retrieval Augmented Generation . https://doi.org/10.2139/ssrn.6101386
Aryani, A. (2024). A Brief Introduction to Retrieval Augmented Generation (RAG). https://doi.org/10.59350/8mwjx-hry76
Budha, K., & Lagun, N. (2026). Operationalizing Reliability Gaps in Large Language Models: A Semi-Systematic Evidence Map of Reasoning, Factuality, Evaluation, and Retrieval-Augmented Generation. https://doi.org/10.21203/rs.3.rs-9882260/v1
Madhurima, M. (2026). FaithGate: A Proposed Architecture for Real-Time Hallucination Detection and Correction in Retrieval-Augmented Generation Systems. https://doi.org/10.2139/ssrn.7242858