MEGA Hub

EnterpriseRAG: Benchmarking LLM Instruction Adherence and Robustness under Non-Ideal Enterprise Retrieval

Authors

Do you know Huiqi Miao?You can claim authorship or link another user.Do you know Xinbao Sun?You can claim authorship or link another user.Do you know Bo Wang?You can claim authorship or link another user.Do you know Fanyu Meng?You can claim authorship or link another user.Do you know Lijun Mei?You can claim authorship or link another user.Do you know Na Wu?You can claim authorship or link another user.Do you know Di Jin?You can claim authorship or link another user.Do you know Chao Deng?You can claim authorship or link another user.Do you know Junlan Feng?You can claim authorship or link another user.

Abstract

Enterprise RAG deployments face a critical reliability gap: while LLMs satisfy 80% of individual constraints, only 26.8% of responses meet all requirements simultaneously, revealing a 57-point orchestration gap. Existing benchmarks assume clean retrieval with simple queries, failing to capture production conditions where noisy documents and multi-dimensional constraints coexist. We introduce EnterpriseRAG, a benchmark of 983 expert-validated samples across six domains that systematically simulates three failure modes absent from prior work: retrieval noise, knowledge gaps, and factual conflicts, coupled with complex instructions. Evaluation of 13 state-of-the-art LLMs reveals a severe instruction adherence collapse, where high per-constraint satisfaction masks low holistic compliance. Critical findings expose deep barriers under knowledge gaps and factual conflicts, even with reasoning-enhanced inference, indicating production RAG requires explicit context-aware protocols and calibrated judgment. EnterpriseRAG provides a reproducible foundation for measuring and closing these gaps, directly informing deployment decisions for enterprise-scale RAG systems. We will release the benchmark and evaluation framework upon publication.

Community

00