mlsanityData to 14 Oct 2026

Position: It is Time to Virtualize Foundation Models with a Self-evolving Operating System Layer

arXiv 2609.19203 ↗First posted 16 Sep 2026cs.AI cs.LG cs.MAcited by 0

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

also in 1 more
Abstract (arXiv)

AI applications have shifted from single, monolithic foundation models (FM) to compound agentic systems. Yet today's stacks remain fragmented: even as protocols (e.g., MCP, A2A) ease tool/agent connectivity, each framework embeds an implicit runtime for state, memory, budgets, and guardrails, making behavior non-portable and governance brittle. It mirrors computing before operating systems, when every program re-implemented basic services. This position paper argues that the field now needs a Foundation Model Operating System (FMOS) -- a system layer that virtualizes FM interactions analogous to how virtual machines abstract physical hardware, giving applications the illusion of dedicated, trustworthy FM instances with effectively unbounded capabilities. Internally, the FMOS orchestrates knowledge across memory tiers, model selection and resource allocation, and verification and policy enforcement. Like the human brain switching between fast intuition and slow deliberation, the FMOS learns when to intervene and when to let inference proceed directly and continuously adapting its policies based on operational experience.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 32 of its 72 references that are in the map. First posted on arXiv 16 Sep 2026.

Builds on

32 of 72 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025DeepAgent: A General Reasoning Agent with Scalable Toolsets24 Oct
  2. 2025EvoTest: Evolutionary Test-Time Learning for Self-Improving Agentic Systems15 Oct
  3. 2025Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems31 Mar
  4. 2025SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planning15 Mar
  5. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains17 Jun

Other references · 27

  1. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,988
  2. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  3. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  4. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,336
  5. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 797
  6. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
  7. 2024The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructionscited by 493
  8. 2025L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learningcited by 384
  9. 2024Improving Alignment and Robustness with Circuit Breakerscited by 322
  10. 2025Chain of Draft: Thinking Faster by Writing Lesscited by 297
  11. 2024Foundation Models for Time Series Analysis: A Tutorial and Surveycited by 239
  12. 2024Llumnix: Dynamic Scheduling for Large Language Model Servingcited by 174
15 more
  1. 2023Large Language Model Routing with Benchmark Datasetscited by 172
  2. 2025Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memorycited by 136
  3. 2024vAttention: Dynamic Memory Management for Serving LLMs without PagedAttentioncited by 98
  4. 2024AIOS: LLM Agent Operating Systemcited by 97
  5. 2025MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scalingcited by 81
  6. 2024From Isolated Conversations to Hierarchical Schemas: Dynamic Tree Memory Representation for LLMscited by 79
  7. 2025When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMscited by 53
  8. 2025Reasoning Language Models: A Blueprintcited by 42
  9. 2025Thinking vs. Doing: Agents that Reason by Scaling Test-Time Interactioncited by 34
  10. 2024InferCept: Efficient Intercept Support for Augmented Large Language Model Inferencecited by 33
  11. 2025Representation Engineering for Large-Language Models: Survey and Research Challengescited by 21
  12. 2024A Blueprint Architecture of Compound AI Systems for Enterprisecited by 6
  13. 2024The Compressor-Retriever Architecture for Language Model OScited by 5
  14. 2025Forecast2Anomaly (F2A): Adapting Multivariate Time Series Foundation Models for Anomaly Predictioncited by 3
  15. 2020DBOS: A Proposal for a Data-Centric Operating Systemcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024A Taxonomy of Architecture Options for Foundation Model-based Agents: Analysis and Decision Model
  2. 2024Foundation Model Sherpas: Guiding Foundation Models through Knowledge and Reasoning
  3. 2026Agentic AIs Are the Missing Paradigm for Out-of-Distribution Generalization in Foundation Models
  4. 2026Agency as a New Organizing Abstraction for Computer Science
  5. 2025Foundation Models as World Models: A Foundational Study in Text-Based GridWorlds
  6. 2024Deploying Foundation Model Powered Agent Services: A Survey
  7. 2026Towards an Agent Operating System - Lessons from Classical and Cloud OS
  8. 2025Single-Agent Scaling Fails Multi-Agent Intelligence: Towards Foundation Models with Native Multi-Agent Intelligence