Context-Fidelity Boosting: Enhancing Faithful Generation through Watermark-Inspired Decoding
On the map: LLM safety and security → Watermarking generated content and models
Abstract (arXiv)
Large language models (LLMs) often produce content that contradicts or overlooks information provided in the input context, a phenomenon known as faithfulness hallucination. In this paper, we propose Context-Fidelity Boosting (CFB), a lightweight and general decoding-time framework that reduces such hallucinations by increasing the generation probability of source-supported tokens. Motivated by logit-shaping principles from watermarking techniques, CFB applies additive token-level logit adjustments based on a token's degree of support from the input context. Specifically, we develop three boosting strategies: static boosting, which applies a fixed bias to source-supported tokens; context-aware boosting, which scales this bias using the divergence between next-token distributions with and without context; and token-aware boosting, which further redistributes the adaptive bias according to local relevance estimated from source-position attention and source-scoped semantic similarity. CFB requires no retraining or architectural changes, making it compatible with a wide range of LLMs. Experiments on summarization and question answering tasks across multiple open-source LLMs show that CFB consistently improves faithfulness metrics with minimal generation overhead. Our implementation is fully open-sourced.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Watermarking generated content and models, one of 12 directions in LLM safety and security. Builds on 26 of its 38 references that are in the map. First posted on arXiv 24 Apr 2026.
Builds on
26 of 38 references.
References · 26
- 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
- 2018Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarizationcited by 1,308
- 2017Get To The Point: Summarization with Pointer-Generator Networkscited by 1,281
- 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,103
- 2021Entity-Based Knowledge Conflicts in Question Answeringcited by 349
- 2023Trusting Your Evidence: Hallucinate Less with Context-aware Decodingcited by 319
- 2023RAGTruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Modelscited by 272
- 2024Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Mapscited by 140
- 2023Chatlaw: A Multi-Agent Legal Assistant based on a Role-Aligned Mixture-of-Experts Architecturecited by 119
- 2024ClashEval: Quantifying the tug-of-war between an LLM's internal prior and external evidencecited by 109
- 2024Summary of a Haystack: A Challenge to Long-Context LLMs and RAG Systemscited by 87
- 2024FaithEval: Can Your Language Model Stay Faithful to Context, Even If "The Moon is Made of Marshmallows"cited by 85
14 more
- 2024A Survey of Large Language Models in Finance (FinLLMs)cited by 77
- 2023FactKB: Generalizable Factuality Evaluation using Language Models Enhanced with Factual Knowledgecited by 55
- 2024AdaCAD: Adaptively Decoding to Balance Conflicts between Contextual and Parametric Knowledgecited by 35
- 2024Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy Constraintcited by 27
- 2024Emotional RAG: Enhancing Role-Playing Agents through Emotional Retrievalcited by 26
- 2024Mitigating Large Language Model Hallucination with Faithful Finetuningcited by 24
- 2024Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs?cited by 24
- 2024DeCoRe: Decoding by Contrasting Retrieval Heads to Mitigate Hallucinationscited by 23
- 2024A Probabilistic Framework for LLM Hallucination Detection via Belief Tree Propagationcited by 21
- 2024Improving Factuality in Large Language Models via Decoding-Time Hallucinatory and Truthful Comparatorscited by 19
- 2024Model Tells Itself Where to Attend: Faithfulness Meets Automatic Attention Steeringcited by 15
- 2024RuleRAG: Rule-Guided Retrieval-Augmented Generation with Language Models for Question Answeringcited by 13
- 2026OpenDecoder: Open Large Language Model Decoding to Incorporate Document Quality in RAGcited by 11
- 2026ReAttn: Improving Attention-based Re-ranking via Attention Re-weightingcited by 5
Built on by
Citing papers per year: 2026: 1.
1 paper.