SOURCE-LINKED INTELLIGENCE

What Eviction Destroys: A Restore-Counterfactual Audit of Forgetting in Agent Memory

arXiv · AI, language, vision and robotics · article · Sep 8, 2026 · UTC

Agent memory systems must discard stored information when their history exceeds a fixed token budget. Existing budget-accuracy frontiers quantify the resulting loss in accuracy, but do not distinguish irreversible losses caused by eviction from recoverable retrieval failures. We introduce the restore counterfactual, a per-question paired intervention that reinstates the question's gold evidence in the read-time context and reruns the same reader. Combining the change in correctness with whether the evidence was retained after eviction classifies each oracle-answerable error as recoverable, irr

Read original source ↗ Open in workspace

recordType

paper

region

Global