Karina· @karinanguyen · X·· 1 小时前AI 评分56
AI 导读
Karina Nguyen 分享一个 Inkling 2 Small 智能体的实测轨迹：当申报工具只返回 PSEG 的目录时，该智能体自行下载完整的 3.7MB 10-Q，用 Python 解析，并发现需从经营现金流中剔除的 12 亿美元抵押品释放，得分 0.82，高于次优模型的 0.67。引用的 @thoughtfullab 内容称其在 DiligenceBench 上表现良好，该模型为 594B 参数、17B 激活参数的 MoE，定价 $0.30/M 输入和 $1/M 输出 token。
正文 · 原文
One of my favorite traces: when the filing tool returned only PSEG’s table of contents, Inkling 2 Small agent independently downloaded the full 3.7MB 10-Q, parsed it in Python, and caught a $1.2B collateral release that needed to be excluded from operating cash flow. It scored 0.82 versus 0.67 for the next-best model.
Congrats @miramurati @johnschulman2 et al on this cool model!
Inkling 2 Small can generate a full public equity diligence memo for just $0.69 using our finance harness. We got early access to the model from @thinkymachines and have been evaluating it on DiligenceBench, where it’s performing remarkably well. It’s a 594B-parameter MoE model with 17B active parameters, priced at $0.30/M input and $1/M output tokens.在 X 查看被引用的帖子
来源：Karina · x.com