Keywords
retrieval self-check
distractor density
paired simulation
reproducibility
Abstract
We evaluated retrieval self-check for retrieval-augmented answering under 25% distractor density. A deterministic paired simulation generated 72 cases and preserved a boundary-condition stratum. Mean evidence faithfulness changed from 0.585 to 0.631; the paired difference was +0.046 (95% interval +0.044 to +0.048). 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
S, S. B., & N, K. (2026). UAMG-RAG: Adaptive Multi-Granularity Retrieval-Augmented Generation with Uncertainty-Guided Re-Retrieval. https://doi.org/10.21203/rs.3.rs-10020935/v1
Srivastava, M. (2026). A Privacy-First Architecture for Fully Local Retrieval-Augmented Generation in Secure Document Intelligence. https://doi.org/10.36227/techrxiv.176800894.46972585/v1
Samal, M. P. (2026). A Theoretical Analysis of Self-Contained Retrieval-Augmented Generation with Small Language Models. https://doi.org/10.36227/techrxiv.177219989.96070478/v1
Procko, T., & Ochoa, O. (2024). Graph Retrieval-Augmented Generation for Large Language Models: A Survey. https://doi.org/10.2139/ssrn.4895062
Roy, P., & Singirikonda, A. (2026). Optimization and Constraint Modeling using LLMs with a Retrieval Augmented Generation Process. https://doi.org/10.2139/ssrn.7068218
Karasan, A. (2026). Improving Retrieval-Augmented Generation: A Systematic Literature Review of Retrieval-Enhancement Techniques. https://doi.org/10.21203/rs.3.rs-10368581/v1
Khan, N., Islam, M. S., & Saha, S. (2026). CyberRerankBench: Benchmarking Rerankers for Cybersecurity Retrieval-Augmented Generation. https://doi.org/10.2139/ssrn.7240159
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
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