
If you're building agents that accumulate large (like trace or log data), this walks through why naive truncation breaks reasoning and summarization cedes too much control, and offers a battle-tested pattern: head/tail preservation with a retrievable memory store plus offloading.
“the best context strategy is one that lets your agents remember what it needs to um, and forget what it doesn't”
Sally-Ann Delucia
“context management isn't is a product and a UX problem, not just an engineering one”
Sally-Ann Delucia
“the main takeaway from this was that over truncation had broke the reasoning. It couldn't remember.”
Sally-Ann Delucia
“We were kind of hoping to get a little bit of a a secret from from them, but I guess we'll all just have to keep you know, doing our own research there.”
Sally-Ann Delucia
“agents don't fail because of prompts, they fail because of context”
Sally-Ann Delucia
videoWhy AI Agents Need Million-Token Context — Thomas Wolf & Olive Song, MiniMax
videoYour company brain will leak secrets: how we stopped it for big banks — Tanmai Gopal, PromptQL
videoTethered: Our Agents Are Us — Shu Fang, Two Sigma
videoAgents' next frontier: agent-to-agent and network effects — Jean-Denis Greze, TownChecking sign-in…
Loading comments…