Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy Constraint

2024/02/19 by Yuan, Xiaowei, Yang, Zhao, Wang, Yequan +3 · 9 citations
#Artificial Intelligence (cs.AI) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2402.11893

Abstract

Large language models internalize enormous parametric knowledge during pre-training. Concurrently, realistic applications necessitate external contextual knowledge to aid models on the underlying tasks. This raises a crucial dilemma known as knowledge conflicts, where the contextual knowledge clashes with the However, existing decoding works are specialized in resolving knowledge conflicts and could inadvertently deteriorate performance in absence of conflicts. In this paper, we propose an adaptive decoding method, termed as contextual information-entropy constraint decoding (COIECD), to discern whether the knowledge conflicts occur and resolve them. It can improve the model's faithfulness to conflicting context, and simultaneously maintain high performance among non- Our experiments show that COIECD exhibits strong performance and robustness over knowledge conflicts in realistic datasets. Code is available.

Cited by

• CoCoA: Confidence and Context-Aware Adaptive Decoding for Resolving Knowledge Conflicts in Large Language Models
• HAVE: Head-Adaptive Gating and ValuE Calibration for Hallucination Mitigation in Large Language Models
• Continuously Steering LLMs Sensitivity to Contextual Knowledge with Proxy Models
• Conflict-Aware Soft Prompting for Retrieval-Augmented Generation
• Deep Research: A Survey of Autonomous Research Agents
• Format as a Prior: Quantifying and Analyzing Bias in LLMs for Heterogeneous Data
• FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation
• Can LLMs Reconcile Knowledge Conflicts in Counterfactual Reasoning
• Micro-Act: Mitigating Knowledge Conflict in LLM-based RAG via Actionable Self-Reasoning

Related

• AdaCAD: Adaptively Decoding to Balance Conflicts between Contextual and Parametric Knowledge
• DYNAMICQA: Tracing Internal Knowledge Conflicts in Language Models
• Entropy Adaptive Decoding: Dynamic Model Switching for Efficient Inference
• Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineering
• Taming Knowledge Conflicts in Language Models