mlsanityData to 7 Oct 2026

CUE-R: Beyond the Final Answer in Retrieval-Augmented Generation

arXiv 2604.05467 ↗First posted 7 Apr 2026cs.IR cs.CL cs.LGcited by 2

On the map: Retrieval-augmented generation → RAG robustness and adaptive retrieval

Abstract (arXiv)

As language models shift from single-shot answer generation toward multi-step reasoning that retrieves and consumes evidence mid-inference, evaluating the role of individual retrieved items becomes more important. Existing RAG evaluation typically targets final-answer quality, citation faithfulness, or answer-level attribution, but none of these directly targets the intervention-based, per-evidence-item utility view we study here. We introduce CUE-R, a lightweight intervention-based framework for measuring per-evidence-item operational utility in single-shot RAG using shallow observable retrieval-use traces. CUE-R perturbs individual evidence items via REMOVE, REPLACE, and DUPLICATE operators, then measures changes along three utility axes (correctness, proxy-based grounding faithfulness, and confidence error) plus a trace-divergence signal. We also outline an operational evidence-role taxonomy for interpreting intervention outcomes. Experiments on HotpotQA and 2WikiMultihopQA with Qwen-3 8B and GPT-5.2 reveal a consistent pattern: REMOVE and REPLACE substantially harm correctness and grounding while producing large trace shifts, whereas DUPLICATE is often answer-redundant yet not fully behaviorally neutral. A zero-retrieval control confirms that these effects arise from degradation of meaningful retrieval. A two-support ablation further shows that multi-hop evidence items can interact non-additively: removing both supports harms performance far more than either single removal. Our results suggest that answer-only evaluation misses important evidence effects and that intervention-based utility analysis is a practical complement for RAG evaluation.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation. Builds on 28 of its 35 references that are in the map. First posted on arXiv 7 Apr 2026.

Builds on

28 of 35 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Evidence Contextualization and Counterfactual Attribution for Conversational QA over Heterogeneous Data with RAG Systems13 Dec
  2. 2024Groundedness in Retrieval-augmented Long-form Generation: An Empirical Study10 Apr
  3. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection17 Oct
  4. 2023Benchmarking Large Language Models in Retrieval-Augmented Generation4 Sep

Other references · 24

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  4. 2025Qwen3 Technical Reportcited by 8,195
  5. 2017On Calibration of Modern Neural Networkscited by 4,860
  6. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  7. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  8. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,126
  9. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  10. 2021Improving language models by retrieving from trillions of tokenscited by 1,278
  11. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,259
  12. 2023AgentBench: Evaluating LLMs as Agentscited by 1,211
12 more
  1. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,081
  2. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 668
  3. 2023Enabling Large Language Models to Generate Text with Citationscited by 584
  4. 2023API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMscited by 574
  5. 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 571
  6. 2023Active Retrieval Augmented Generationcited by 506
  7. 2023Adaptive Chameleon or Stubborn Sloth: Revealing the Behavior of Large Language Models in Knowledge Conflictscited by 333
  8. 2022Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoningcited by 331
  9. 2024Evaluation of Retrieval-Augmented Generation: A Surveycited by 146
  10. 2022Attributed Question Answering: Evaluation and Modeling for Attributed Large Language Modelscited by 108
  11. 2024RAGChecker: A Fine-grained Framework for Diagnosing Retrieval-Augmented Generationcited by 59
  12. 2024Correctness is not Faithfulness in RAG Attributionscited by 18

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026EMBER: Efficient Memory via Budgeted Evidence Retention for Long-Horizon Agentscited by 5
  2. 2026What Would Fix This RAG Failure? Auditing Counterfactual Response with Paired Evidence Interventionscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Assessing the Downstream Utility of Evidence-Aware Retrieval in RAG
  2. 2026The RAT: A Unified Bayesian Model for RAG Evaluation
  3. 2026Why Retrieval-Augmented Generation Fails: A Graph Perspective
  4. 2026Relevance Is Not Sufficient Evidence: Detecting Evidence Gaps Before Generation in RAG
  5. 2025Worse than Zero-shot? A Fact-Checking Dataset for Evaluating the Robustness of RAG Against Misleading Retrievals
  6. 2026SURE-RAG: Sufficiency and Uncertainty-Aware Evidence Verification for Selective Retrieval-Augmented Generation
  7. 2025Towards Transparent RAG: Fostering Evidence Traceability in LLM Generation via Reinforcement Learning
  8. 2026When Confidence Takes the Wrong Path: Diagnosing Retrieval-State Lock-In in RAG