Catalog Systems Recent method papers — theorized, no distinct product UFO2

UFO2

https://arxiv.org/abs/2504.14603

Desktop AgentOS with memory-driven task automation.

Active · Verified 2026-10-28 · fresh as of 2026-10-28 · 18 cited facts, 0 with a third-party source · Suggest a correction

At a glance

Type
Desktop AgentOS vendor fix
Tier
T4
Section
Recent method papers — theorized, no distinct product
Created
2025-04 vendor fix
Latest release
not applicable — not OSS fix
License
not applicable — not OSS fix
GitHub
not applicable — no GitHub repo fix
Pricing
not applicable — not commercial fix
Funding
not applicable — not commercial fix
Freshness
2026-10-28 fix
Layer
agent-infra repo fix

Memory taxonomy

storage
hybrid
retrieval
similarity
persistence
cross-session
update
extraction
unit
skill
governance
user-controllable
conflict
llm-arbitrate

When to use

Optimised for: not applicable - research paper

Anti-fit: not applicable - research paper

Pros & cons

Pros

User-focused agent with memory of preferences and patterns.

Cons

Preprint; user-modeling quality varies by domain.

Claims & capabilities

Multiagent operating system for Windows with HostAgent task coordination and specialized AppAgents using native APIs; hybrid control detection fusing Windows UIA with vision parsing; speculative multi-action planning reduces per-step LLM overhead; PiP interface; substantial improvements in robustness and execution accuracy over prior CUAs across 20+ real-world Windows applications

Technical surface

API surface
not applicable — research paper
Backend storage
not applicable — research paper
Deployment
not applicable — not a deployable product
Embedding model
not applicable — research paper
Multi-tenancy
not applicable — research paper
MCP
not applicable — research paper, no deployed product
A2A
not applicable — research paper, no deployed product
OpenTelemetry
not applicable — research paper, no deployed product

Similar systems

Other recent method papers — theorized, no distinct product in the catalog, ranked by inbound references.

  • Compressive Transformer T3

    Maintains recent states in full resolution while compressing older memories with learned compression functions. DeepMind.

  • MemGPT v2 / agent-tools T3

    Already in catalog as the foundational MemGPT paper. Note: Letta is the productionised successor (cross-listed).

  • Transformer-XL T3

    Extends context through segment-level recurrence + caching of hidden states from prior segments. Foundational long-context architecture.

  • Generative Agents T3

    Park et al. — landmark agent-simulation paper. Reflection + memory stream + retrieval enable believable agent behavior.

  • MemoryBank T3

    Enhances LLMs with long-term memory. Early influential paper.

  • Reflexion T3

    Language agents with verbal reinforcement learning.

Sources (17 cited facts, 0 third-party)

Each fact links to the page it came from; the badge says who published it (how to read sources).