3 comments

  • SwellJoe6 minutes ago
    Everybody builds one. And, then they usually figure out that making the model fill its context with a bunch of memories hurts performance more often than it helps.
  • georgespencer36 minutes ago
    Given the abundance of vaguely similar local-first AI memory layers, it might be a good idea to add a "Why Mnemo" section right at the top of README.md to explain why folks should consider using it.
  • bilbo-b-baggins28 minutes ago
    You forgot BM25 embeddings.<p><a href="https:&#x2F;&#x2F;github.com&#x2F;MikeS071&#x2F;ai-engram" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;MikeS071&#x2F;ai-engram</a><p><a href="https:&#x2F;&#x2F;github.com&#x2F;lamost423&#x2F;openclaw-hybrid-memory" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;lamost423&#x2F;openclaw-hybrid-memory</a><p><a href="https:&#x2F;&#x2F;medium.com&#x2F;@qdrddr&#x2F;agentic-memory-framework-hindsight-with-fts-bm25-hybrid-search-and-rabitq-ce8141a0323e" rel="nofollow">https:&#x2F;&#x2F;medium.com&#x2F;@qdrddr&#x2F;agentic-memory-framework-hindsigh...</a><p><a href="https:&#x2F;&#x2F;clawhub.ai&#x2F;vnesin-sarai&#x2F;hybrid-retrieval" rel="nofollow">https:&#x2F;&#x2F;clawhub.ai&#x2F;vnesin-sarai&#x2F;hybrid-retrieval</a><p><a href="https:&#x2F;&#x2F;www.josecasanova.com&#x2F;blog&#x2F;openclaw-qmd-memory" rel="nofollow">https:&#x2F;&#x2F;www.josecasanova.com&#x2F;blog&#x2F;openclaw-qmd-memory</a><p><a href="https:&#x2F;&#x2F;medium.com&#x2F;@richardhightower&#x2F;stop-the-hallucinations-hybrid-retrieval-with-bm25-pgvector-embedding-rerank-llm-rubric-rerank-895d8f7c7242" rel="nofollow">https:&#x2F;&#x2F;medium.com&#x2F;@richardhightower&#x2F;stop-the-hallucinations...</a><p><a href="https:&#x2F;&#x2F;github.com&#x2F;oomkapwn&#x2F;enquire-mcp#-why-its-the-best" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;oomkapwn&#x2F;enquire-mcp#-why-its-the-best</a><p><a href="https:&#x2F;&#x2F;github.com&#x2F;rohitg00&#x2F;agentmemory#key-capabilities" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;rohitg00&#x2F;agentmemory#key-capabilities</a><p><a href="https:&#x2F;&#x2F;github.com&#x2F;Melody-0321&#x2F;NE-Memory-Core" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;Melody-0321&#x2F;NE-Memory-Core</a><p><a href="https:&#x2F;&#x2F;github.com&#x2F;ClaudioDrews&#x2F;memory-os" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ClaudioDrews&#x2F;memory-os</a><p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Okapi_BM25" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Okapi_BM25</a><p>&gt; It is based on the probabilistic retrieval framework developed in the 1970s and 1980s<p>Anyway, good for ya, hope you had fun building it.
    • asdev5 minutes ago
      I haven&#x27;t seen one unique product in AI, everyone is building the same thing