想快速了解新模型水平？Arena 用真实用户偏好做评估，比传统基准快几个数量级，看榜就行。
Arena.ai 今日发布 AutoEval，一种基于奖励模型的新评估方法，该模型由数百万条真实 Arena 用户偏好数据校准。AutoEval 与实时人工评估结果高度一致，能把模型评估耗时从数天缩短到数小时。它目前支持 Text、Vision、Image、Code 四个 Arena 榜单。新模型上线后，AutoEval 会直接在排行榜上给出估算分数。
Learn more about AutoEval and our methodology at: https://t.co/2kEg4iZUSG
Learn more about AutoEval and our methodology at: x.com/arena/status/2… Arena.ai @arena Today we’re launching AutoEval: a new evaluation methodology that ranks models using reward models based on millions of real Arena user preferences. Highlights: - High-quality evaluation signals calibrated on real preference data - Strong alignment with live human evaluations - Evaluations that are several orders of magnitude faster (hours instead of days) - Support for Text, Vision, Image, and Code Arena AutoEval enables us to evaluate newly launched models much faster and share results with the community sooner. We’ll now show AutoEval estimated scores for new models directly on the leaderboard. More details in the thread. 🧵 🔗 View Quoted Tweet 💬 0 🔄 1 ❤️ 1 👀 1159 ⚡
- 向阳乔木07-30 17:43原文