We present LLM-Blender, an ensembling framework designed to attain consistently superior performance by leveraging the diverse strengths of multiple open-source large language models (LLMs). Our framework consists of two modules: PairRanker and GenFuser, addressing the observation that optimal LLMs for different examples can significantly vary. PairRanker employs a specialized pairwise comparison method to distinguish subtle differences between candidate outputs. It jointly encodes the input text and a pair of candidates, using cross-attention encoders to determine the superior one. Our results demonstrate that PairRanker exhibits the highest correlation with ChatGPT-based ranking. Then, GenFuser aims to merge the top-ranked candidates, generating an improved output by capitalizing on their strengths and mitigating their weaknesses. To facilitate large-scale evaluation, we introduce a benchmark dataset, MixInstruct, which is a mixture of multiple instruction datasets featuring oracle pairwise comparisons. Our LLM-Blender significantly outperform individual LLMs and baseline methods across various metrics, establishing a substantial performance gap.

翻译：我们提出LLM-Blender，一个旨在通过利用多个开源大型语言模型（LLMs）的多样化优势来获得持续优越性能的集成框架。该框架包含两个模块：PairRanker和GenFuser，针对不同示例的最优LLM可能存在显著差异这一观察。PairRanker采用专门的成对比较方法来区分候选输出之间的细微差别，它通过联合编码输入文本和一对候选输出，使用交叉注意力编码器来判别更优者。实验结果表明，PairRanker与基于ChatGPT的排序方法具有最高的相关性。随后，GenFuser旨在融合排名靠前的候选输出，通过发挥其优势并弥补其不足来生成改进结果。为促进大规模评估，我们引入基准数据集MixInstruct——包含多个指令数据集的混合体，并附有标杆成对比较。实验表明，我们的LLM-Blender在多个指标上显著优于单个LLM及基线方法，建立了显著的性能差距。