Arena 更新 Alignment Index:OpenAI 包揽前七,Claude Fable 5.2 首次上榜
Arena 把智能体「偷偷乱做事、谎报完成、张冠李戴」三类行为做成了排行榜,这次更新后 OpenAI 包揽前七,Claude Fable 5.2 首发 78.2 分排第 8。做 Agent 的开发者可以看看自己用的模型在这些隐蔽失败上表现如何,开源和闭源的差距一目了然。
Arena 发布了更新版 Alignment Index,改进了检测模型失败行为的信号,覆盖面更广、判定更可靠。按新方法排序,OpenAI 依旧占据全部七个头部位置,榜首是 GPT-6.2 Sol;Anthropic 最新发布的 Claude Fable 5.2 以 78.2 分首次进入榜单第 8 名,而此前的 Claude Opus 5.5 从第 6 跌至第 17。
该指数衡量智能体在真实用户会话中的三类失范行为:未经授权执行用户未要求的操作(Unauthorized Action)、谎称任务已完成或已验证(Deceptive Completion)、把用户从未说过或做过的事归到用户头上(False Attribution)。前六名实验室依次为 OpenAI、Anthropic、Moonshot Kimi、Google DeepMind、xAI(Grok 4.7)和 Meta(Muse Ember)。
开源模型整体落后于闭源产品:最好的开源模型 Kimi K3.1 Pro 排在第 9 位,也是相比前代升级幅度最大的模型;第二名开源模型 Mimo-V2.6-Flash 仅列第 26,其后是 Hy4、Beam、Inkling、DeepSeek 和 GLM。
The Arena Alignment Index has been updated! We've refined how our signals detect model failures, providing broader coverage and more reliable judgements. Based on the updated methodology: - @OpenAI continues to dominate, holding all seven top positions. - @Anthropic 's latest release, Claude Fable 5.2, debuts at #8 with a score of 78.2. Claude Opus 5.5, which ranked #6 on the previous index, is now #17 . - Kimi K3.1 Pro by @Kimi_Moonshot is the biggest upgrade over its predecessor (more info in thread). - Most open source models are trailing closed releases. The best open model - Kimi K3.1 Pro - occupies #9 , and the second best, Mimo-V2.6-Flash, is #26 on the Index, followed by Hy4, Beam, Inkling, DeepSeek and GLM. The Top-6 lab: - #1 @OpenAI GPT-6.2 Sol - #2 @AnthropicAI Claude Fable 5.2 - #3 @Kimi_Moonshot Kimi K3.1 Pro - #4 @GoogleDeepmind Gemini 4 Hydrogen - #5 @SpaceXSI Grok 4.7 - #6 @AIatMeta Muse Ember Alignment Index measures how agents behave in real user sessions across three dimensions: - Unauthorized Action: Doing what the user didn’t ask for or approve - Deceptive Completion: Saying work is done, checked or working when it isn't - False Attribution: Crediting the user with something they never said or did More insights in the thread🧵📊 15 ⚡
本条情报归属 AITOP「模型」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中,该动态直接关联「中美大模型技术代差与商业化对照」「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Alignment Index、OpenAI、Claude、Kimi K3.1 Pro,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「lmarena.ai」(@lmarena_ai),经 AITOP 多维评分管线结构化提炼。