Researchers have introduced a new standard called RTEB (Retrieval-oriented Evaluation Benchmark) to evaluate the performance of retrieval-based models. This standard aims to provide a more comprehensive and accurate way to measure the effectiveness of these models, which are used in applications such as question answering and document search. The RTEB standard includes metrics for evaluating both recall and precision, and it is designed to be more robust than existing evaluation methods. --- Why it matters: This matters because retrieval-based models are widely used in AI applications, and a standardized evaluation method can help improve their performance and reliability. By using the RTEB standard, developers can get a more accurate picture of how well their models work and make targeted improvements. Source: https://huggingface.co/blog/rteb