mlsanityData to 14 Oct 2026

Conflict-Aware Soft Prompting for Retrieval-Augmented Generation

arXiv 2508.15253 ↗First posted 21 Aug 2025cs.CL cs.AIcited by 5

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

Abstract (arXiv)

Retrieval-augmented generation (RAG) enhances the capabilities of large language models (LLMs) by incorporating external knowledge into their input prompts. However, when the retrieved context contradicts the LLM's parametric knowledge, it often fails to resolve the conflict between incorrect external context and correct parametric knowledge, known as context-memory conflict. To tackle this problem, we introduce Conflict-Aware REtrieval-Augmented Generation (CARE), consisting of a context assessor and a base LLM. The context assessor encodes compact memory token embeddings from raw context tokens. Through grounded/adversarial soft prompting, the context assessor is trained to discern unreliable context and capture a guidance signal that directs reasoning toward the more reliable knowledge source. Extensive experiments show that CARE effectively mitigates context-memory conflicts, leading to an average performance gain of 5.0% on QA and fact-checking benchmarks, establishing a promising direction for trustworthy and adaptive RAG systems.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation. Builds on 25 of its 46 references that are in the map. First posted on arXiv 21 Aug 2025.

Builds on

25 of 46 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Enhancing Noise Robustness of Retrieval-Augmented Language Models with Adaptive Adversarial Training31 May
  2. 2024Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity21 Mar
  3. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection17 Oct

Other references · 22

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,673
  3. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,975
  4. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  5. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,212
  6. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 974
  7. 2023An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuningcited by 639
  8. 2021ColBERTv2: Effective and Efficient Retrieval via Lightweight Late Interactioncited by 476
  9. 2023Learning to Compress Prompts with Gist Tokenscited by 390
  10. 2023Making Retrieval-Augmented Language Models Robust to Irrelevant Contextcited by 369
  11. 2023Adapting Language Models to Compress Contextscited by 356
  12. 2023Trusting Your Evidence: Hallucinate Less with Context-aware Decodingcited by 327
10 more
  1. 2024Knowledge Conflicts for LLMs: A Surveycited by 283
  2. 2024ProSA: Assessing and Understanding the Prompt Sensitivity of LLMscited by 149
  3. 2023Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentationcited by 147
  4. 2024xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Tokencited by 125
  5. 2023FactKG: Fact Verification via Reasoning on Knowledge Graphscited by 57
  6. 2024Enhancing Contextual Understanding in Large Language Models through Contrastive Decodingcited by 49
  7. 2024Probing-RAG: Self-Probing to Guide Language Models in Selective Document Retrievalcited by 29
  8. 2024Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy Constraintcited by 27
  9. 2023Nugget: Neural Agglomerative Embeddings of Textcited by 19
  10. 2024From Reading to Compressing: Exploring the Multi-document Reader for Prompt Compressioncited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 4.

5 papers.

  1. 2025Memory in the Age of AI Agentscited by 303
  2. 2026COMI: Coarse-to-fine Context Compression via Marginal Information Gaincited by 13
  3. 2026From Conflict to Consensus: Boosting Medical Reasoning via Multi-Round Agentic RAGcited by 7
  4. 2026Retrieval-Augmented Generation Must Move Beyond Factual Grounding to Represent Diverse Opinionscited by 1
  5. 2026Parameters vs. Context: TRACE Fine-Tuning for Robust Retrieval-Augmented Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Retrieval-Augmented Generation: A Comprehensive Survey of Architectures, Enhancements, and Robustness Frontiers
  2. 2025Grounding Long-Context Reasoning with Contextual Normalization for Retrieval-Augmented Generation
  3. 2026Seeing through the Conflict: Transparent Knowledge Conflict Handling in Retrieval-Augmented Generation
  4. 2024Long-Context LLMs Meet RAG: Overcoming Challenges for Long Inputs in RAG
  5. 2024SFR-RAG: Towards Contextually Faithful LLMs
  6. 2025Oreo: A Plug-in Context Reconstructor to Enhance Retrieval-Augmented Generation
  7. 2024The Power of Noise: Redefining Retrieval for RAG Systems
  8. 2024From RAGs to rich parameters: Probing how language models utilize external knowledge over parametric information for factual queries