mello-uses-contriever-embedding
IN premise — summaries/2026/08/24/zhong-2023-mquake-s5-mello-a-proposal-for-editing-large.md
Created 2026-08-25T02:59:06+00:00
MeLLo converts edited facts to sentence statements via manual templates, embeds them using the pretrained Contriever model (Izacard et al., 2021), and stores embeddings in a retrieval index.
Summary
MeLLo turns each fact edit it makes into a plain sentence using hand-written templates, then converts that sentence into a numeric fingerprint with the Contriever model so it can be quickly looked up later. This means its ability to remember and retrieve past edits depends entirely on the quality of those manual templates and the Contriever embedding, not on any adaptive or learned encoding of the edit's meaning.