AI/ML●●Solid
Breathe-Memory – Associative memory injection for LLMs (not RAG)
Graph-based context compression beats lossy summarization when tokens run out.
Big BrainNiche Gem
mvyshnyvetska
612mo ago
Back to browse
GitHub Repository
An associative memory you can run anywhere. Write facts in plain language, recall them by meaning. No tables, no schema, no embeddings, no model required.
2 starsRust
LLM Memory Solved?
by gary23w·Jun 14, 2026·1 point·0 comments
AI Analysis
●●SolidBig BrainWizardry
130× more facts per GiB than vectors, but no semantic similarity matching.
Strengths
- •Stem-based indexing uses 48 bytes per fact vs 1536-dim float vectors
- •Pure Rust with zero dependencies compiles to 1 MB WebAssembly
- •PlasticNeuron tier implements Hebbian learning with O(1) scalar updates
Weaknesses
- •Cannot match semantically similar queries without embeddings
- •LLM Memory Solved title overclaims what cue-based recall delivers
Category
Target Audience
Developers building local-first AI applications needing lightweight memory
Similar To
Chroma · LanceDB · Mem0
Similar Projects
AI/ML●●●Banger
Mnemora – Serverless memory DB for AI agents (no LLM in your CRUD path)
Removes LLM from memory CRUD path—sub-10ms reads beat Mem0/Zep's 200-500ms by design.
Big BrainSolve My ProblemWizardry
isaacgbc
443mo ago
AI/ML●●Solid
Anchor Engine – Deterministic Semantic Memory for LLMs local (<3GB RAM)
Compresses 28M tokens to 100k queryable chars local-only; duplicates RAG problems at smaller scale.
Solve My ProblemBig Brain
BERTmackliin
213mo ago
AI/ML●●●Banger
Memory system for AI agents with associations, forgetting, synthesis
Biologically-inspired memory consolidation that prunes unused facts and strengthens associations overnight.
Big BrainNiche Gem
jari_mustonen
201mo ago
AI/ML●●Solid
AgentKV – SQLite for AI agent memory (MMAP vector+graph DB)
Single-file mmap storage plus an HNSW vector index and explicit graph edges is an elegant, practical combo — think "SQLite for agent memory" with CRC-32 crash recovery and zero-server convenience. The C++20 core + nanobind gives zero-copy NumPy views and GIL-free searches, and the claimed FAISS-like throughput makes this genuinely interesting for local setups; main caveat is build/toolchain friction and how rich the surrounding ecosystem becomes.
WizardryNiche Gem
shiwang_khera
103mo ago
AI/ML●●●Banger
A memory database that forgets, consolidates, and detects contradiction
Vector DBs store memories; this one forgets, consolidates, and flags contradictions like human memory.
Big BrainZero to OneWizardry
pranabsarkar
48332mo ago