用多个不同模型协作检测并减少幻觉，效果更优且更省算力。
Teaming LLMs to Detect and Mitigate Hallucinations
- 通过整合多个不同来源的LLM响应，提升幻觉检测能力。
- 相比单模型方法，检测准确率显著提升，推理成本反而降低。
- 适合需要高可靠性生成的场景，如医疗、法律等关键领域。
近期研究显示，基于一致性的方法——通过聚合单个LLM对同一提示的多次采样响应——在大语言模型幻觉检测与缓解方面达到领先水平。这类方法可部分弥补训练数据不完善带来的缺陷，如偏见和部署时所需信息的不足。本文表明，将此类单模型一致性方法扩展至融合来自15个不同LLM的响应，这些模型具有不同的训练数据、训练方案和架构，能带来显著超越单模型一致性的幻觉检测与缓解性能提升。我们在多个由15个模型组成的模型团队中评估了这一‘联盟一致性’方法，并探究了在何种条件下联合不同模型更为有益。此外，我们发现这些性能提升往往伴随推理成本的降低，有效缓解了单模型一致性方法的一大缺点。
原文摘要 · Abstract (English)
Recent work has demonstrated state-of-the-art results in large language model (LLM) hallucination detection and mitigation through consistency-based approaches which involve aggregating multiple responses sampled from a single LLM for a given prompt. These approaches help offset limitations stemming from the imperfect data on which LLMs are trained, which includes biases and under-representation of information required at deployment time among other limitations which can lead to hallucinations. We show that extending these single-model consistency methods to combine responses from multiple LLMs with different training data, training schemes and model architectures can result in substantial further improvements in hallucination detection and mitigation capabilities beyond their single-model consistency counterparts. We evaluate this "consortium consistency" approach across many model teams from a pool of 15 LLMs and explore under what conditions it is beneficial to team together different LLMs in this manner. Further, we show that these performance improvements often come with reduced inference costs, offsetting a significant drawback with single-model consistency methods.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。