GA

Gabe Pereyra

精选演讲 主持
本站收录 1 集 · 13 条金句 · 关联 10 个

① 他说过的话

13 条

如果你是一家应用层公司,与前沿实验室竞争是一场不公平的游戏。有富有的团队,有贫穷的团队,然后还有我们这些应用层的。
And it’s an unfair game competing with the Frontier Labs if you’re an application layer company. There are rich teams, there are poor teams, and then there’s us in the application layer.
—— Gabe Pereyra · [00:51]

指向原始笔记的链接

但今天,利用前沿生态系统,我认为你可以与前沿实验室竞争,并构建前沿智能。
But today, using the Frontier ecosystem, I think you can compete with the Frontier Labs and build Frontier Intelligence.
—— Gabe Pereyra · [01:37]

指向原始笔记的链接

如果你没有一个好的基准测试,你就无法训练模型。而如果你不能训练模型,你就不需要在生产环境中部署它们。
If you don’t have a good benchmark, you can’t train models. And if you can’t train models, you don’t need to serve them in production.
—— Gabe Pereyra · [02:07]

指向原始笔记的链接

我们与最大的律师事务所和企业合作,他们的法律数据极其敏感,受特权保护。所以你不能把它放进通用模型里——我们甚至不能把它放进我们自己的模型里。
We work with the largest law firms and enterprises, and their legal data is incredibly sensitive. It’s privileged. And so you can’t put it in generic models. We can’t even put it in our models.
—— Gabe Pereyra · [03:21]

指向原始笔记的链接

就像现在的工程师不再写代码,而是 vibe code 并引导这些编码模型一样,我们也开始做同样的事情。
the same way that engineers now don’t write code, they vibe code and guide these coding models, we’re starting to do the same thing.
—— Gabe Pereyra · [03:48]

指向原始笔记的链接

过去做后训练是不值得的,因为模型通过预训练提升得太快,你做的任何后训练很快就会被下一个预训练模型吸收掉。
In the past, it wasn’t worth doing post-training because the models were improving so quickly from pre-training that any post-training you did quickly got absorbed by the next pre-trained model.
—— Gabe Pereyra · [06:04]

指向原始笔记的链接

也许不是通用的前沿智能,但如果你有一个具体的任务,比如像我们这样,它们是有竞争力的。
Maybe not general frontier intelligence, but if you have a specific task like us, They are competitive.
—— Gabe Pereyra · [06:29]

指向原始笔记的链接

通常,如果你和他们合作却没能得到更好的结果,那很可能是你的数据集有问题。
And usually, if you work with them and you’re not able to get better results, there’s probably something you’re doing wrong with your data set.
—— Gabe Pereyra · [06:50]

指向原始笔记的链接

然后我们称之为产品反馈。我称之为愤怒的客户邮件。
And then we call it product feedback. I call it angry customer emails.
—— Gabe Pereyra · [11:40]

指向原始笔记的链接

我认为我们选择法律领域的原因之一是,如果你只是构建一个这些顶级律所正在处理的真实客户案件,前沿模型仍然做不了这个。
And I think part of why we picked the legal domain is if you just build a realistic client matter that these top firms are working on, like the frontier model still can’t do this.
—— Gabe Pereyra · [19:19]

指向原始笔记的链接

你可以有一个在我们基准测试上表现非常好的模型,但它们在某种程度上过拟合了,然后你把它放进一个更通用的助手类产品里,当你超出分布范围时它就有点崩溃了。
you can have a model that does very well on our benchmarks but they’ve overfit to some degree and then you put it in a more generic assistant like product and it kind of falls apart for when you go out of distribution
—— Gabe Pereyra · [22:29]

指向原始笔记的链接

但我认为找到某种形式的持续学习——这里的终局不是让我们构建最好的法律模型,而是帮助每家律所或企业客户把它定制成他们所做工作的类型。
But I think figuring out some form of continual learning is I think the endgame here is not for us to build the best legal model. It’s for us to help every law firm or enterprise customer customize it to the type of work they’re doing.
—— Gabe Pereyra · [26:15]

指向原始笔记的链接

就是你如何以一种水平产品不会去做的方式,极度垂直地深入你自己的领域?
is how do you just go hyper-vertical into your domain in a way that the horizontal products won’t?
—— Gabe Pereyra · [29:05]

指向原始笔记的链接

② 出现在这些集

1 集

③ 他谈到的

点进去有真内容 —— 本页主要出口

Harvey · 后训练 · 基准测试 · 合成数据 · 开源模型 · 智能体 · RL 环境 · 模型路由 · 长上下文 · DeepMind

④ 也在聊「智能体」的人