Putting the Reasoning Trace Before the Context Instead of After Lifts GraphWalks Exact Match From 29% to 82%

arXiv 2609.02702·high signalXBlueskyLinkedInCopy link

Trace as State formalizes a mismatch in causal transformers: long-context reasoning often depends on task state discovered only later, and for causal state-update processors, giving the condition first can require exponentially less memory in the worst case than giving it last. The method reuses collected reasoning traces as a textual proxy for task state and places them before the long-context block on a fresh pass. Against Trace Append, a matched control placing the identical trace after the context, it won in 26 of 27 model/task/metric combinations; on GraphWalks Parents exact match went from 29.2% (initial pass) and 43.0% (Trace Append) to 81.8% for DeepSeek V4 Pro Preview, and from 66.4% and 83.2% to 100.0% for GLM-5.2.