AITOPAI热报
How-to

Optiver 工程师分享：如何在用户提问前预载 200K token 上下文

AI Engineer@aiDotEngineer
Post
View the post
Why we picked it

Optiver 的工程师讲怎么用 tool search 和缓存把 200K token 上下文压下去，做智能体的可以看看思路。

Optiver 的 Noyan Tokgozoglu 介绍了在用户提问前就构建好 200K token 上下文的做法。分享核心是用 tool search 削减上下文膨胀，配合缓存和更好的默认设置来控制成本。这是为 11 月 10 日至 12 日在旧金山举办的 AIE CODE 大会做的预热内容。

Image · AI Engineer
Original · AI Engineer

200K tokens before the user asks a question. Noyan Tokgozoglu of @OptiverGlobal on cutting context bloat with tool search, caching and better defaults. Bring your agent stack to AIE CODE. Nov 10 to 12, SF. Apply: ai.engineer/code Your browser does not support the video tag. 🔗 View on Twitter 💬 0 🔄 0 ❤️ 4 👀 680 📊 1 ⚡

#tool search#上下文缓存#智能体#AIE CODE#Optiver

Keep reading

Recent stories with the same tag or section
How-toBuildersSelected02:50
LangChain explains common agent retrieval failures and fixes

Most agent failures blamed on model quality are actually retrieval design problems, and this framework gives builders a checklist to diagnose which failure mode they are seeing. Developers building agent harnesses or RAG pipelines can use the guidance to improve search precision, trim returned content and keep sources fresh. LangChain's own tooling is referenced, so the patterns map directly to its ecosystem.

#LangChain#RAG#agents
官方账号LangChain@LangChainAISource