大模型
x.com ·
@Arena.ai (@arena)
Arena 分析 30,086 场对战：模型平均仅共享 43% 观点
Arena 对 30,086 场 Battle Mode 对战进行分析，发现模型平均只共享 43% 的观点。同实验室或同国家的模型并未表现出一致的更高概念重叠，Claude Fable 5 最接近的概念匹配既非 Opus 也非 Sonnet。完整发现见其外链文章。
AI推荐理由 Arena 官方账号发布的模型响应相似度分析预告，给出 30,086 场对战、平均共享 43% 观点等具体数字，并指出同实验室/同国家模型未必更相似、Claude Fable 5 最接近的概念匹配并非 Opus 或 Sonnet。但正文仅为引子，核心结论需跳转外链文章，信息密度有限，且未提供方法细节与可验证数据，故质量分中等。
62
浏览 78,397
点赞 534
评论 41
转发 43
正文
We analyzed how similar model responses were across 30,086 Arena battles.
Models shared 43% of their ideas on average. We might expect that models from the same lab, or country, would show greater conceptual overlap. But the results don’t consistently support that.
Claude Fable 5 illustrates this pattern: its closest conceptual match was neither Opus nor Sonnet. Which model came closest, along with the broader findings, may surprise you.
Check out the full article from @DawidGalarowicz and @petergostev below.