llm-kb
← Back to research
Paper

Shared Selective Persistent Memory Architecture Proposed for Agentic LLMs

A new arxiv paper introduces shared selective persistent memory, an architecture that identifies and retains four categories of reusable context (task specifications, data schemas, tool configurations, and tool-use patterns) for agentic LLM systems. This approach aims to solve the context problem where each session starts from zero, while avoiding token-inefficient naive persistence of entire conversation histories. The work is relevant to developers building multi-turn code generation agents.

0 engagement·1 source·Fri, Jul 10, 2026, 03:07 PM
The paper, posted on arxiv on 2026-07-10, addresses a fundamental limitation in agentic LLM systems that generate code through multi-turn tool use: each session discards configuration choices, domain constraints, data schemas, and tool-use patterns from previous sessions. The proposed architecture selectively persists reusable context across sessions, improving generation quality and token efficiency. No specific model names, parameter counts, or benchmark numbers are provided in the post.

Entities

arxiv(tool)Shared Selective Persistent Memory(concept)

Related

ProductSun, Jul 12, 2026, 10:28 PM

Collaborative context-sharing memory platform for AI agents and teams

A platform that enables AI agents and human teams to share and persist context across sessions. It uses LLMs to manage memory, allowing agents to recall past interactions and collaborate on tasks. Solves the problem of fragmented context in multi-agent systems and team workflows.

2 engagement·1 source·hackernews
ProductSun, Jul 12, 2026, 09:08 PM

Persistent memory for AI assistants via MCP

Adaptive Recall provides persistent memory for AI assistants using the Model Context Protocol (MCP). It solves the problem of AI assistants forgetting context between sessions, enabling long-term personalized interactions.

1 engagement·1 source·hackernews
PaperFri, Jul 10, 2026, 08:10 AM

GRACE: Graph-Regularized Agentic Context Evolution for Reliable Long-Horizon LLM Agents

A new arXiv paper proposes GRACE, a method that maintains persistent system-level instructions for LLM agents as a typed semantic graph instead of flat text. This graph-regularized approach enables scoped verification and reliable context evolution over long horizons under distribution shift, addressing verification difficulties from accumulated instructions.

0 engagement·1 source·arxiv
arXiv
ProductSat, Jul 11, 2026, 08:01 PM

odek: AI agent with long-term semantic memory across sessions

odek is an AI agent that maintains a structured, semantic long-term memory across sessions, remembering user preferences, codebases, and goals without requiring re-explanation. It uses LLMs to process and store information in a three-tier memory system, solving the problem of session-only memory for users who need continuity in their interactions.

0 engagement·1 source·rss
RSS
CommunitySat, Jul 11, 2026, 01:56 PM

Developer reports AI coding agent with persistent memory across cold reboots

A developer on Reddit reports that their AI coding agent retained full context—including decisions, boundaries, and past mistakes—across a complete PC shutdown and fresh terminal session. The agent continued mid-thought without re-explanation or warm-up, suggesting a breakthrough in long-term memory persistence for coding assistants.

3 engagement·1 source·reddit