Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SuperLocalMemory V3.3: The Living Brain -- Biologically-Inspired Forgetting, Cognitive Quantization, and Multi-Channel Retrieval for Zero-LLM Agent Memory Systems

Varun Pratap Bhardwaj

Abstract

AI coding agents operate in a paradox: they possess vast parametric knowledge yet cannot remember a conversation from an hour ago. Existing memory systems store text in vector databases with single-channel retrieval, require cloud LLMs for core operations, and implement none of the cognitive processes that make human memory effective. We present SuperLocalMemory V3.3 ("The Living Brain"), a local-first agent memory system implementing the full cognitive memory taxonomy with mathematical lifecycle dynamics. Building on the information-geometric foundations of V3.2 (arXiv:2603.14588), we introduce five contributions: (1) Fisher-Rao Quantization-Aware Distance (FRQAD) -- a new metric on the Gaussian statistical manifold achieving 100% precision at preferring high-fidelity embeddings over quantized ones (vs 85.6% for cosine), with zero prior art; (2) Ebbinghaus Adaptive Forgetting with lifecycle-aware quantization -- the first mathematical forgetting curve in local agent memory coupled to progressive embedding compression, achieving 6.7x discriminative power; (3) 7-channel cognitive retrieval spanning semantic, keyword, entity graph, temporal, spreading activation, consolidation, and Hopfield associative channels, achieving 70.4% on LoCoMo in zero-LLM Mode A; (4) memory parameterization implementing Long-Term Implicit memory via soft prompts; (5) zero-friction auto-cognitive pipeline automating the complete memory lifecycle. On LoCoMo, V3.3 achieves 70.4% in Mode A (zero-LLM), with +23.8pp on multi-hop and +12.7pp on adversarial. V3.2 achieved 74.8% Mode A and 87.7% Mode C; the 4.4pp gap reflects a deliberate architectural trade-off. SLM V3.3 is open source under the Elastic License 2.0, runs entirely on CPU, with over 5,000 monthly downloads.

SuperLocalMemory V3.3: The Living Brain -- Biologically-Inspired Forgetting, Cognitive Quantization, and Multi-Channel Retrieval for Zero-LLM Agent Memory Systems

Abstract

AI coding agents operate in a paradox: they possess vast parametric knowledge yet cannot remember a conversation from an hour ago. Existing memory systems store text in vector databases with single-channel retrieval, require cloud LLMs for core operations, and implement none of the cognitive processes that make human memory effective. We present SuperLocalMemory V3.3 ("The Living Brain"), a local-first agent memory system implementing the full cognitive memory taxonomy with mathematical lifecycle dynamics. Building on the information-geometric foundations of V3.2 (arXiv:2603.14588), we introduce five contributions: (1) Fisher-Rao Quantization-Aware Distance (FRQAD) -- a new metric on the Gaussian statistical manifold achieving 100% precision at preferring high-fidelity embeddings over quantized ones (vs 85.6% for cosine), with zero prior art; (2) Ebbinghaus Adaptive Forgetting with lifecycle-aware quantization -- the first mathematical forgetting curve in local agent memory coupled to progressive embedding compression, achieving 6.7x discriminative power; (3) 7-channel cognitive retrieval spanning semantic, keyword, entity graph, temporal, spreading activation, consolidation, and Hopfield associative channels, achieving 70.4% on LoCoMo in zero-LLM Mode A; (4) memory parameterization implementing Long-Term Implicit memory via soft prompts; (5) zero-friction auto-cognitive pipeline automating the complete memory lifecycle. On LoCoMo, V3.3 achieves 70.4% in Mode A (zero-LLM), with +23.8pp on multi-hop and +12.7pp on adversarial. V3.2 achieved 74.8% Mode A and 87.7% Mode C; the 4.4pp gap reflects a deliberate architectural trade-off. SLM V3.3 is open source under the Elastic License 2.0, runs entirely on CPU, with over 5,000 monthly downloads.

Paper Structure

This paper contains 51 sections, 3 theorems, 13 equations, 4 figures, 9 tables.

Table of Contents

  1. Introduction
  2. The Session Amnesia Problem
  3. The Cognitive Memory Gap
  4. Contributions
  5. Relationship to Prior Work
  6. Paper Organization
  7. Background and Related Work
  8. Agent Memory Systems
  9. Cognitive Memory Architectures
  10. Vector Compression and Quantization
  11. Memory and Forgetting in AI
  12. Spreading Activation and Graph Memory
  13. Memory Parameterization
  14. System Architecture
  15. Architecture Overview
  16. ...and 36 more sections

Key Result

Theorem 4.1

For $b$-bit TurboQuant applied to any unit-norm vector $\mathbf{x} \in \mathbb{R}^d$: This is within $2.7\times$ of the information-theoretic lower bound. $\blacktriangleleft$$\blacktriangleleft$

Figures (4)

  • Figure 1: SLM V3.3 system architecture. The Interface Layer provides 60 MCP tools, a CLI with daemon serve mode (32$\times$ cold-start speedup), a 17-tab web dashboard, and auto-cognitive hooks for Claude Code. The Engine Layer implements 7-channel cognitive retrieval, Ebbinghaus lifecycle management with EAP precision scheduling, FRQAD/TurboQuant quantization (C1), soft prompt generation (C4), and a code knowledge graph module. The Storage Layer uses local SQLite databases with sqlite-vec for vector operations. Orange blocks indicate novel contributions.
  • Figure 2: Mixed-precision preference: percentage of 18,840 query-fact pairs where the f32 embedding is correctly preferred over the 4-bit quantized version. FRQAD achieves perfect precision (100%) by accounting for quantization uncertainty via variance inflation on the Fisher-Rao geodesic.
  • Figure 3: Ebbinghaus retention curves over 30 simulated days. Hot facts (daily access) converge toward the polar4 tier ($R \approx 0.35$). Warm facts (every 3 days) show a characteristic cyclic pattern. Cold facts decay immediately below the forget threshold. Dotted lines indicate EAP precision tier boundaries.
  • Figure 4: LoCoMo per-category comparison. V3.3 R3 surpasses Paper 2 on adversarial (+6.1pp) and substantially closes the multi-hop gap (+23.8pp from baseline). The single-hop regression ($-$14.9pp vs Paper 2) reflects 7-channel fusion complexity.

Theorems & Definitions (6)

  • Theorem 4.1: MSE Distortion Upper Bound turboquant
  • Definition 4.2: FRQAD
  • Proposition 4.3: Monotonic degradation
  • Definition 5.1: Memory Strength
  • Definition 5.2: Retention
  • Theorem 5.3: Convergence of Ebbinghaus-Fokker-Planck System