Lune

ACL Findings2023

RHO: Reducing Hallucination in Open-domain Dialogues with Knowledge Grounding

Ziwei Ji, Zihan Liu, Nayeon Lee, Tiezheng Yu, Bryan Wilie, Min Zeng, Pascale Fung

Abstract

Dialogue systems can leverage large pretrained language models and knowledge to generate fluent and informative responses. However, these models are still prone to produce hallucinated responses not supported by the input source, which greatly hinders their application. The heterogeneity between external knowledge and dialogue context challenges representation learning and source integration, which further contributes to unfaithfulness. To handle this challenge and generate more faithful responses, this paper presents RHO (ρ) utilizing the representations of linked entities and relation predicates from a knowledge graph (KG). We propose (1) local knowledge grounding to combine textual embeddings with the corresponding KG embeddings; and (2) global knowledge grounding to equip RHO with multi-hop reasoning abilities via the attention mechanism. In addition, we devise a response re-ranking technique based on walks over KG sub-graphs for better conversational reasoning. Experimental results on OpenDialKG (Moon et al., 2019) show that our approach significantly outperforms stateof-the-art methods on both automatic and human evaluation by a large margin, especially in hallucination reduction (17.54% in FeQA (Durmus et al., 2020)). 1

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

13 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • KCTS: Knowledge-Constrained Tree Search Decoding with Token-Level Hallucination Detection cites Plug and Play Language Models: A Simple Approach to Controlled Text Generation
  • KCTS: Knowledge-Constrained Tree Search Decoding with Token-Level Hallucination Detection cites Retrieval Augmentation Reduces Hallucination in Conversation
  • Making Large Language Models Perform Better in Knowledge Graph Completion cites Language Models are Few-Shot Learners
  • Retrieval Augmentation Reduces Hallucination in Conversation cites Language Models are Few-Shot Learners
  • KG-BART: Knowledge Graph-Augmented BART for Generative Commonsense Reasoning cites K-BERT: Enabling Language Representation with Knowledge Graph
  • KG-BART: Knowledge Graph-Augmented BART for Generative Commonsense Reasoning cites Language Models are Few-Shot Learners
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers13

  1. Mitigating Large Language Model Hallucinations via Autonomous Knowledge Graph-Based RetrofittingXinyan Guan, Yanjiang Liu, Hongyu Lin, Yaojie Lu et al.AAAI 2024 · 127 citations
  2. Making Large Language Models Perform Better in Knowledge Graph CompletionYichi Zhang, Zhuo Chen, Lingbing Guo, Yajing Xu et al.ACM MM 2024 · 86 citations
  3. Teaching Language Models to Hallucinate Less with Synthetic TasksErik Jones, Hamid Palangi, Clarisse Simões, Varun Chandrasekaran et al.ICLR 2024 · 43 citations
  4. ANAH-v2: Scaling Analytical Hallucination Annotation of Large Language ModelsYuzhe Gu, Ziwei Ji, Wenwei Zhang, Chengqi Lyu et al.NeurIPS 2024 · 20 citations
  5. KCTS: Knowledge-Constrained Tree Search Decoding with Token-Level Hallucination DetectionSehyun Choi, Tianqing Fang, Zhaowei Wang, Yangqiu SongEMNLP 2023 · 10 citations

Builds on18

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. Plug and Play Language Models: A Simple Approach to Controlled Text GenerationSumanth Dathathri, Andrea Madotto, Janice Lan, Jane Hung et al.ICLR 2020 · 1,166 citations
  3. K-BERT: Enabling Language Representation with Knowledge GraphWeijie Liu, Peng Zhou, Zhe Zhao, Zhiruo Wang et al.AAAI 2020 · 898 citations
  4. Retrieval Augmentation Reduces Hallucination in ConversationKurt Shuster, Spencer Poff, Moya Chen, Douwe Kiela et al.EMNLP Findings 2021 · 579 citations
  5. KG-BART: Knowledge Graph-Augmented BART for Generative Commonsense ReasoningYe Liu, Yao Wan, Lifang He, Hao Peng et al.AAAI 2021 · 220 citations

Related papers

  1. Neural Path Hunter: Reducing Hallucination in Dialogue Systems via Path GroundingNouha Dziri, Andrea Madotto, Osmar Zaïane, Avishek Joey BoseEMNLP 2021 · 74 citations
  2. Paths-over-Graph: Knowledge Graph Empowered Large Language Model ReasoningXingyu Tan, Xiaoyang Wang, Qing Liu, Xiwei Xu et al.WWW 2025 · 86 citations
  3. Post-hoc Utterance Refining Method by Entity Mining for Faithful Knowledge Grounded ConversationsYoonna Jang, Suhyune Son, Jeongwoo Lee, Junyoung Son et al.EMNLP 2023
  4. AttnIO: Knowledge Graph Exploration with In-and-Out Attention Flow for Knowledge-Grounded DialogueJaehun Jung, Bokyung Son, Sungwon LyuEMNLP 2020 · 33 citations
  5. KGA-LM: Representation-Level Grounding for Conversational Search over Knowledge GraphsYunfei Li, Chengfei Liu, Rui Zhou, Zhiyu Xu et al.KDD 2026
Dusk over the sea between two cliffs drawn in fine vertical lines