A Practical Guide to Memory for Autonomous LLM Agents

Quoted on this wiki

Every place a page here uses this source, in the order the words come in it.

1. The paper leads with an empirical observation that should recalibrate your priorities if it hasn't already: “The gap between 'has memory' and 'does not have memory' is often larger than the gap between different LLM backbones.” This is a huge claim. Swapping your underlying model matters less than whether your agent can remember things. I've felt this intuitively, but seeing it stated this plainly in a formal survey is useful. Practitioners spend enormous energy on model selection and prompt tuning while treating memory as an afterthought. That's backward.

   In AI SDR agent context and memory

2. The taxonomy maps surprisingly well to what I built in OpenClaw through iteration and frustration. That's either validating or humbling, depending on how you look at it (probably both.) The paper formalizes patterns that practitioners have been discovering independently, which is what a good survey should do. “Evaluation is still primitive.” Memory is where the real differentiation happens in agent systems. Not the model, not the prompts. The memory architecture. The paper gives you a vocabulary and a framework to think more clearly about it.

   In AI SDR agent evaluation

Quotes are verbatim and belong to their publishers. Where a quote has been translated, the original wording is kept below it.