Why front-loaded rules drift in long evaluator and agent loops

Subtitle: In context is not the same as in control. By Beamlak Adane A colleague asked a sharp question that many of us hit once we move from demos to production evaluators and agents: In multi-turn evaluation or agent loops, models often begin to ignore the initial rules in the system prompt even though those tokens are still inside the context window. What is happening at the token level during decoding that causes early “anchor” tokens to lose influence over time in a streaming context, and h