AI agents have an optimal effective context length where intelligence peaks
David Hershey · How Claude Plays Pokémon was made · Mar 4, 2025 · at 18:57
Anthropic engineer David Hershey reflects on tuning conversation history length for agent harnesses on Latent Space.
0:00 / 0:19exact quote · 19.4s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“I think like one thing you see a lot when you talk to people building agents is there's like some effective context length that actually like has the model be the smartest. And that seems to vary slightly model by model, but for this model, for whatever purpose, like this, the 30 message worked better than 20 and better than 40.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from David Hershey
Hershey: Claude 3.7 extended thinking does not help Pokémon gameplay
“I've tested, like, all sorts of the extended thinking mode with, ah, 3.7 on it, and, like, it doesn't really help.”
Hershey: Anthropic Study Found Claude Treats Named Characters Better
“Anthropic actually did like a blinded study of like named characters versus unnamed characters in different settings, and Claude like actually does clearly prefer and is nicer to named characters, which is an interesting thing.”
Upgrading Claude models in Pokémon primarily involves deleting prompt scaffolding
“Literally every model that has come out with Pokemon, like, the main change that I have made to this agent is deleting prompt stuff.”
Prompting Claude to nickname its Pokémon caused it to exhibit protective behaviors
“And one thing we found when we started doing that is it got more protective of the Pokemon it nicknamed. Like, it's pretty obvious, like, when it catches a Pokemon, Now that it has a nickname, it will, like, go heal it right away if it's hurt, and that did not…”
Claude 3.7 Sonnet uniquely generates meta-commentary reflecting on its perceptual errors
“And actually like one of the things that's most unique about. 3.7 sonnet that I've seen is like, it will have like meta commentary on what it's good at and bad at and it's knowledge base. Like I misperceived this thing. And so like, I need to be careful doing …”