LAM is a lossy framework with three components: a deterministic deduplication rule that bounds retrieval-score perturbation, a memory manager that preserves the cached prefix and overlaps compaction with inference, and a performance model that estimates compaction costs. On 600 agent trajectories, it removed 22.47% of observation tokens while retaining 99.984% of the measured gold-patch evidence. The performance model predicts a 71.4x to 91.6x end-to-end speedup when deletions happen before prefill instead of after, and this benefit comes from the scheduling rather than the rule itself.
LAM Cuts Agent Memory While Bounding Retrieval Score Errors
A deterministic compaction method removed redundant observations before model prefill while preserving nearly all measured evidence across 600 agent trajectories.
Argonne National Laboratory · University of Houston · Oregon State University · IBM Research
In one line
LAM trims 22.47% of agent memory tokens with 99.984% evidence retention and predicts up to 91.6x speedup by scheduling compaction before prefill.
Before you read this
Three ideas this paper builds onLAM combines three components: a deterministic deduplication rule with a bound on how much substitution can perturb retrieval scores, a memory manager that preserves the cached prefix and overlaps compaction with inference, and a performance model for estimating compaction costs.
How it works
before and afterThe Daily Edition
The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.