AI 精选动态
智能评分 60
Agents-A1 35B模型得分79.0超更大模型
AI 推荐理由
该数据点来自尚未正式发布的论文，值得关注Agents-A1架构细节。核心解读
一篇论文声称其Agents-A1（35B模型）在FrontierScience-Olympiad上得分为79.0，高于Kimi-K2.6（73.0）和DeepSeek-V4-pro（76.0）。
全文
one of the paper’s strongest claims: Agents-A1 (a 35B model) scores 79.0 on FrontierScience-Olympiad, beating larger rivals like Kimi-K2.6 at 73.0 and DeepSeek-V4-pro at 76.0.
FrontierScience-Olympiad evaluates olympiad-level scientific reasoning across fields like physics, chemistry, and biology.
