mlsanityData to 7 Oct 2026

From Facts to Conclusions : Integrating Deductive Reasoning in Retrieval-Augmented LLMs

arXiv 2512.16795 ↗First posted 18 Dec 2025cs.CL cs.AI cs.CYcited by 0

On the map: LLM reasoning → Knowledge- and tool-augmented reasoning

also in 1 more
Abstract (arXiv)

Retrieval-Augmented Generation (RAG) grounds large language models (LLMs) in external evidence, but fails when retrieved sources conflict or contain outdated or subjective information. Prior work address these issues independently but lack unified reasoning supervision. We propose a reasoning-trace-augmented RAG framework that adds structured, interpretable reasoning across three stages : (1) document-level adjudication, (2) conflict analysis, and (3) grounded synthesis, producing citation-linked answers or justified refusals. A Conflict-Aware Trust-Score (CATS) pipeline is introduced which evaluates groundedness, factual correctness, refusal accuracy, and conflict-behavior alignment using an LLM-as-a-Judge. Our 539-query reasoning dataset and evaluation pipeline establish a foundation for conflict-aware, interpretable RAG systems. Experimental results demonstrate substantial gains over baselines, most notably with Qwen, where Supervised Fine-Tuning improved End-to-End answer correctness from 0.069 to 0.883 and behavioral adherence from 0.074 to 0.722.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning. Builds on 15 of its 26 references that are in the map. First posted on arXiv 18 Dec 2025.

Builds on

15 of 26 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Enhancing Noise Robustness of Retrieval-Augmented Language Models with Adaptive Adversarial Training31 May
  2. 2023Retrieval-Augmented Generation for Large Language Models: A Survey18 Dec

Other references · 13

  1. 2025Qwen3 Technical Reportcited by 8,195
  2. 2024GPT-4o System Cardcited by 4,467
  3. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,355
  4. 2023Mistral 7Bcited by 3,065
  5. 2021Improving language models by retrieving from trillions of tokenscited by 1,278
  6. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
  7. 2025SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-trainingcited by 655
  8. 2024A Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Modelscited by 450
  9. 2021SituatedQA: Incorporating Extra-Linguistic Contexts into QAcited by 145
  10. 2024ConflictBank: A Benchmark for Evaluating the Influence of Knowledge Conflicts in LLMcited by 73
  11. 2024Measuring and Enhancing Trustworthiness of LLMs in RAG through Grounded Attributions and Learning to Refusecited by 33
  12. 2024Open Domain Question Answering with Conflicting Contextscited by 18
1 more
  1. 2024ALiiCE: Evaluating Positional Fine-grained Citation Generationcited by 15

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Exploring Knowledge Conflicts for Faithful LLM Reasoning: Benchmark and Method
  2. 2025Rethinking All Evidence: Enhancing Trustworthy Retrieval-Augmented Generation via Conflict-Driven Summarization
  3. 2025Retrieval is Not Enough: Enhancing RAG Reasoning through Test-Time Critique and Optimization
  4. 2024RAG-Star: Enhancing Deliberative Reasoning with Retrieval Augmented Verification and Refinement
  5. 2026Beyond Factual Accuracy: Evaluating Global Reasoning Integrity in RAG Systems with LogicScore
  6. 2025Synergizing RAG and Reasoning: A Systematic Review
  7. 2025RAG+: Enhancing Retrieval-Augmented Generation with Application-Aware Reasoning
  8. 2025Fishing for Answers: Exploring One-shot vs. Iterative Retrieval Strategies for Retrieval Augmented Generation