📄 中文摘要
谷歌正式发布了 Android Bench，这是一个专门用于评估大型语言模型（LLMs）在安卓开发任务中表现的新排行榜和评估框架。该框架的相关数据集、方法论和测试工具已开源，并可在 GitHub 上公开获取。传统的编码基准测试往往无法全面反映 LLM 在特定开发环境中的性能，因此 Android Bench 旨在填补这一空白，通过系统化的评估方法和任务设计，为开发者提供更准确的性能指标和比较依据。
谷歌人工智能发布 Android Bench：针对安卓开发的 LLM 评估框架和排行榜
📄 English Summary
Google AI Releases Android Bench: An Evaluation Framework and Leaderboard for LLMs in Android Development
Google has officially launched Android Bench, a new leaderboard and evaluation framework aimed at assessing the performance of Large Language Models (LLMs) specifically in Android development tasks. The framework includes an open-source dataset, methodology, and test harness, all of which are publicly available on GitHub. Traditional coding benchmarks often fail to adequately capture the performance of LLMs in specific development environments. Android Bench addresses this gap by providing a systematic evaluation methodology and task design, offering developers more accurate performance metrics and comparative insights.