context window

Definition

How much text a model can pay attention to at one time.

The maximum sequence length a transformer can attend over during a forward pass, bounded by model architecture and memory.

Also called: context windows

Mentioned in 14 episodes

1. 051
   Why Parallel Sampling Plateaus, And What Evidence Graphs Do Instead
2. 049
   An AI Agent Reached for Root in Twelve Minutes, Without Being Attacked
3. 044
   How One Sentence and a Forged History Flip the Most Aligned Models
4. 030
   Why Your AI Agent Won't Stop Working — and Each Model Falls for a Different Trap
5. 028
   Teaching a Model to Hire Copies of Itself: Recursive Agent Optimization
6. 027
   When AI Agents Build the Serving Stack: A Bet on Bespoke Infrastructure
7. 024
   An AI Agent That Found 28 Zero-Days in Windows — And What Made It Work
8. 022
   Training the Model Spec Directly: An Alignment Lever Aimed at the Say-Do Gap
9. 016
   Why Your Coding Agent Stalls While the GPU Runs Hot
10. 014
    Why a Constrained Pipeline Beat a Full Coding Agent at Finding Bugs 30-to-1
11. 012
    Why AI Coding Agents Keep Trying to Debug Without a Debugger
12. 005
    Why a Debugger Designed for Humans Is the Wrong Tool for an AI Agent
13. 003
    How to Pick the Best of Sixteen Coding Agent Rollouts
14. 002
    An AI Ran a Real Optics Lab for 21 Hours and Found a Transformer-Shaped Pattern in Light

Related concepts