Is Factuality Decoding a Free Lunch for LLMs? Evaluation on Knowledge Editing Benchmark

arXiv:2404.00216v1 Announce Type: new
Abstract: The rapid development of large language models (LLMs) enables them to convey factual knowledge in a more human-like fashion. Extensive efforts have been made to reduce factual hallucinations by modifying LLMs with factuality decoding. However, they also pose risks of hindering knowledge updates, as they make models overly confident in known facts. In this work, we first revisite the current factuality decoding methods and verified their effectiveness in enhancing factual accuracy. Subsequently, we conduct further …

abstract arxiv benchmark cs.ai cs.cl decoding development editing evaluation fashion free hallucinations however human human-like knowledge language language models large language large language models llms reduce risks them type