Artificial Analysis· @ArtificialAnlys · X·· 1 小时前AI 评分40
AI 导读
Nemotron 3.5 Super VL 在 Intelligence Index 每项任务上使用 46k 输出 token,比其前代 Nemotron 3 Super(114k)少约 60%,也少于 K2 Horizon 375B A23B(52k)和 MiMo-V2.6-Pro(64k) 尽管 token 效率明显提升,Nemotron 3.5 Super VL 在 Intelligence Index vs. 每任务输出 Token 前沿对比中仍落后不少
正文
Nemotron 3.5 Super VL uses 46k output tokens per Intelligence Index task, ~60% fewer than its predecessor Nemotron 3 Super (114k) and fewer than K2 Horizon 375B A23B (52k) and MiMo-V2.6-Pro (64k)
Despite a clear increase in token efficiency, Nemotron 3.5 Super VL still falls a way behind the Intelligence Index vs. Output Tokens per Task frontier
来源:Artificial Analysis · x.com