LoCoMo — Hebb Mind vs Letta

System	Score	Source
Hebb Mind	95.75% R@10 (bge-large + rerank) / 94.14% (bge-large default) / 91.41% (MiniLM-384), full 1,978q each	LoCoMo
Letta	TBD	No first-party LoCoMo result we could find in their public repo

Why this row is TBD

Letta (formerly MemGPT) does not publish a first-party LoCoMo result on their main repo or blog as of this writing. Third-party benchmarks have appeared but use ad-hoc judges and scenario counts.

To publish a same-row comparison, we would need to run Letta through the Hebb Mind eval/ harness; see vs mem0 for the same caveats.

Open a PR if Letta has since published a LoCoMo number that should be referenced here.