A visual explainer of MIA’s core claim: combine non-parametric memory (compressed external search experience) with parametric memory (planner skill in weights) so a research agent reuses prior investigations without drowning in raw trajectories.
Three specialized roles split the workload: the Manager curates compressed experience, the Planner turns question + memory into strategy, and the Executor interacts with tools and evidence.
Hover boxes and links for state load, update frequency, and auditability tradeoffs.
Existing agent memory often accumulates long episodes verbatim. MIA’s pitch is to retain planning-relevant structure while reducing retrieval clutter and storage burden.
A conceptual schema: not every thought survives, but task decomposition, productive tools, evidence paths, and failure recovery do.
The episode highlights a tension: the paper seems strongest on benchmark performance and weaker on hard operational efficiency curves such as store growth, latency, and context cost.
The core loop is retrieval → planning → execution → reflection → judgment → update. This tab shows the progressive migration between external episodes and planner skill.
Conceptual trend: explicit memory dominates early, then planner internalization rises while external memory remains a searchable safety net.
Compact map of the cited ideas around memory, reflection, RAG, long-horizon agents, and test-time adaptation.