Agent / long-term memory
Memora
Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity
Superseded baseline#32 of 63 most-superseded · first seen Feb 3, 2026
Superseded — cited as a baseline and beaten by newer methods
1 papers critique it · 1 beat it on benchmarks
What papers say
Verbatim critique sentences, each from a paper that cites Memora as a baseline.
In each case, the LLM does the heavy lifting of deciding what is worth remembering and how to organize it---a form of pre-query intelligence that is invisible in per-query token counts.
Beaten on benchmarks
Head-to-head results where a newer method reports beating Memora. Values are copied from the source paper's tables — verify against the cited paper.
SmartSearch beats Memora
91.9 vs 86.3
Overall · [MemOS protocol (gpt-4o-mini)]
SmartSearch: How Ranking Beats Structure for Conversational Memory Retrieval
What to use instead
Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.
- May 11, 2026
- Mar 16, 2026