[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82531-en":3,"doc-seo-82531-105":30,"detail-sidebar-cat-0-en-105":83},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},82531,549758146520,"Patrick","https://ap-avatar.wpscdn.com/avatar/80002397d8c0411e94?_k=1775819394049821470",8,"Research & Report","Dual-Confidence Contrastive Decoding for Retrieval-Augmented Generation","Retrieval-augmented generation (RAG) must answer questions using multiple retrieved documents, yet only some sources are relevant while evidence may be stale, noisy, or mutually inconsistent. This work targets intra-context conflict within the retrieved bundle, proposing DRQA, a factual-conflict QA benchmark built from synthetic enterprise-specific facts. It also introduces DCCD, a training-free decoding method combining document-level and token-level confidence to guide confidence-gated contrastive generation, improving multi-document QA performance, especially on DRQA.","Dual-Confidence Contrastive Decoding for Retrieval-Augmented  \nGeneration  \nRaymond Li♠♢ Md Tawkat Islam Khondaker♠♢ Amirhossein Abaskohi♢ Gabriel Murray♢ Giuseppe Carenini♢ Issam H. Laradji♠♢  \n♠ ServiceNow Research ♢University of British Columbia  \narXiv :2607 .00570v 1 [ cs .CL] 1 Jul 2026  \nAbstract  \nRetrieval-augmented generation (RAG) increasingly requires models to answer questions from multiple retrieved documents, where only some sources are relevant and the retrieved bundle may contain stale, noisy, or conflicting evidence. Existing contrastive decoding methods primarily focus on resolving conflicts between the model’s internal memory and the retrieved context. In contrast, we study the complementary problem of intra-context conflict in multi-document RAG. To evaluate this setting, we introduce DRQA, a factual-conflict question answering benchmark derived from enterprise deep-research scenarios, where answers are grounded in synthetic enterprise-specific facts that are designed not to be recoverable from the model’s internal memory. We further propose Dual-Confidence Contrastive Decoding (DCCD), a training-free decoding method that combines document-level confidence, which estimates whether a document appears sufficient for answering the question, with token-level confidence, which estimates whether that document supports a confident next-token prediction. DCCD selects positive and negative document-conditioned streams using these dual-confidence signals and scales a document-level contrast by their confidence margin. Across DRQA and standard multidocument QA benchmarks, DCCD achieves the best average performance among full-context and contrastive decoding baselines, with the largest gains on DRQA. These results highlight the importance of source-aware, confidencegated decoding when retrieved evidence is internally conflicting.  \n1 Introduction  \nRetrieval-augmented generation (RAG) has become an essential paradigm for grounding large language models (LLMs) in external knowledge (Gao et al., 2023), enabling them to answer domainspecific and knowledge-intensive queries beyond  \nwhat is stored in their parameters. Initially developed for open-domain question answering (QA)(Karpukhin et al., 2020 ; Lewis et al., 2020), RAG has become an integral component of workflows that require resolving information from multiple sources, from open-domain deep-research agents that synthesize evidence across the web (Du et al., 2026 ; Java et al., 2026) to enterprise systems that retrieve from heterogeneous data such as reports, emails, and chats (Choubey et al., 2025 ; Abaskohiet al., 2026) .  \nDespite these advancements, a central challenge for RAG systems is the resolution of conflicts among the knowledge sources (Xu et al., 2024) . At inference time, the LLM can be overly reliant on its internal memory acquired during pre-training, even when the retrieved context provides newer or more task-specific evidence (Chen et al., 2022 ; Xie et al., 2024) . One popular training-free approach to context-memory conflict is contrastive decoding (CD) (Li et al., 2023), which modulates generation by contrasting model predictions under different conditioning contexts (Shi et al., 2024 ; Wang et al., 2025a ; Khandelwal et al., 2025) . Rather than finetuning the LLMs, contrastive decoding adjusts token probabilities based on the difference between output distributions with and without the context. However, this formulation treats the retrieved evidence as a single context and does not resolve conflicts within the retrieved context itself.  \nIn multi-document QA, the model needs to decide the relevance of each document to the query and which evidence to use (Wan et al., 2024 ; Jin et al., 2024b) . The retrieved evidence may contain answer-supporting evidence alongside irrelevant distractors, stale information, and plausible but incorrect claims. In this setting, the decoding problem is not only about choosing the next token, but also about choosin","cbCaioaqFINVDxIG","https://ap.wps.com/l/cbCaioaqFINVDxIG","pdf",576343,3,1,23,"English","en",105,"# Introduction\n## Retrieval-augmented generation and conflicts\n## Contrastive decoding and its limitation\n## Intra-context conflict in multi-document RAG","[{\"question\":\"How does DCCD combine document-level and token-level confidence?\",\"answer\":\"DCCD uses document-level confidence to estimate whether a document is sufficient and token-level confidence to estimate whether that document supports confident next-token prediction; it then selects document-conditioned streams and scales the document-level contrast by their confidence margin.\"}]",1784181294,58,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":78,"head_meta":80,"extra_data":82,"updated_unix":28},"dual-confidence-contrastive-decoding-for-retrieval-augmented-generation","",{"@graph":36,"@context":77},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":20},"https://docshare.wps.com/document/research-report/",{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/dual-confidence-contrastive-decoding-for-retrieval-augmented-generation/82531/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-20","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71],{"name":72,"@type":73,"acceptedAnswer":74},"How does DCCD combine document-level and token-level confidence?","Question",{"text":75,"@type":76},"DCCD uses document-level confidence to estimate whether a document is sufficient and token-level confidence to estimate whether that document supports confident next-token prediction; it then selects document-conditioned streams and scales the document-level contrast by their confidence margin.","Answer","https://schema.org",{"og:url":51,"og:type":79,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":81,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":84},[85,89,93,97,102,107,112,115,120,123,127],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":86,"show_sort_weight":87,"slug":88},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":90,"show_sort_weight":91,"slug":92},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Exam",70,"exam",{"id":98,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},5,"Comic",60,"comic",{"id":103,"doc_module":4,"doc_module_name":46,"category_name":104,"show_sort_weight":105,"slug":106},6,"Technology",50,"technology",{"id":108,"doc_module":4,"doc_module_name":46,"category_name":109,"show_sort_weight":110,"slug":111},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":113,"slug":114},30,"research-report",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},9,"Religion & Spirituality",20,"religion-spirituality",{"id":118,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":118,"slug":122},"World Cup","world-cup",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":124,"slug":126},10,"Lifestyle","lifestyle",{"id":128,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":98,"slug":130},19,"General","general"]