GRAD: Graph-Retrieved Adaptive Decoding for Hallucination Mitigation
By: Manh Nguyen , Sunil Gupta , Dai Do and more
Potential Business Impact:
Makes AI tell the truth, not make things up.
Hallucination mitigation remains a persistent challenge for large language models (LLMs), even as model scales grow. Existing approaches often rely on external knowledge sources, such as structured databases or knowledge graphs, accessed through prompting or retrieval. However, prompt-based grounding is fragile and domain-sensitive, while symbolic knowledge integration incurs heavy retrieval and formatting costs. Motivated by knowledge graphs, we introduce Graph-Retrieved Adaptive Decoding (GRAD), a decoding-time method that grounds generation in corpus-derived evidence without retraining. GRAD constructs a sparse token transition graph by accumulating next-token logits across a small retrieved corpus in a single forward pass. During decoding, graph-retrieved logits are max-normalized and adaptively fused with model logits to favor high-evidence continuations while preserving fluency. Across three models and a range of question-answering benchmarks spanning intrinsic, extrinsic hallucination, and factuality tasks, GRAD consistently surpasses baselines, achieving up to 9.7$\%$ higher intrinsic accuracy, 8.6$\%$ lower hallucination rates, and 6.9$\%$ greater correctness compared to greedy decoding, while attaining the highest truth--informativeness product score among all methods. GRAD offers a lightweight, plug-and-play alternative to contrastive decoding and knowledge graph augmentation, demonstrating that statistical evidence from corpus-level token transitions can effectively steer generation toward more truthful and verifiable outputs.
Similar Papers
Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment
Computation and Language
Makes AI understand facts better, stops fake answers.
ProgRAG: Hallucination-Resistant Progressive Retrieval and Reasoning over Knowledge Graphs
Artificial Intelligence
Helps computers answer hard questions using smart steps.
GRIL: Knowledge Graph Retrieval-Integrated Learning with Large Language Models
Machine Learning (CS)
Helps AI answer questions by learning from connected facts.