SEAgent
https://arxiv.org/abs/2508.04700
Self-evolving computer-use agent with autonomous learning.
Active · Verified 2026-10-28 · fresh as of 2026-10-28 · 19 cited facts, 0 with a third-party source · Suggest a correction
At a glance
- Type
- Self-evolving computer-use agent vendor fix
- Tier
- T4
- 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
- vector
- retrieval
- similarity
- persistence
- long-term
- update
- agent-controlled
- unit
- skill
- governance
- inspectable
- conflict
- llm-arbitrate
When to use
Optimised for: not applicable - research paper
Anti-fit: not applicable - research paper
Pros & cons
Pros
Self-evolving agent with memory of past architectures.
Cons
Preprint; self-evolution stability concerns.
Claims & capabilities
Computer use agent that autonomously evolves through software interactions; combines World State Model for trajectory assessment, Curriculum Generator for progressive tasks, and adversarial imitation+GRPO learning; 23.2% improvement (11.3%->34.5%) over UI-TARS open-source CUA across five novel software environments in OS-World; specialist-to-generalist consolidation surpasses ensembles
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 (18 cited facts, 0 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is Self-evolving computer-use agent vendor
- What it does & what's distinct Self-evolving computer-use agent with autonomous learning. vendor
- Claims / benefits Computer use agent that autonomously evolves through software interactions; combines Wor… vendor
- Modalities text + image vendor
- Created 2025-08 vendor
- Mindshare GitHub: github.com/SunzeY/SEAgent; star count not retrieved repo
- Citations 30 cites 30/yr paper
- Founders / pedigree no data — searched arxiv.org/abs/2508.04700; author affiliations not found in preview vendor
- Memory volume / scale no scale claim — research vendor
- Performance 23.2% absolute improvement (11.3% -> 34.5%) over baseline computer-use agents via World … vendor
- Reproducibility Too recent vendor
- Schema evolution schema with migration vendor
- Contradiction handling LLM resolves and overwrites vendor
- Pros Self-evolving agent with memory of past architectures. vendor
- Cons Preprint; self-evolution stability concerns. vendor
- Project & sources arxiv 2508.04700 vendor
- citation-trajectory [{"year":2025,"cum":1,"infl":0}] paper
- taxonomy-three-layer-rollup agent-infra repo