MemGen
https://arxiv.org/abs/2509.24704
Weaves generative latent memory for self-evolving agents.
Active · Verified 2026-10-28 · fresh as of 2026-10-28 · 21 cited facts, 0 with a third-party source · Suggest a correction
At a glance
- Type
- Generative latent memory 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
- parametric
- retrieval
- parametric-recall
- persistence
- parametric-permanent
- update
- parametric-edit
- unit
- weight
- governance
- opaque
- conflict
- llm-arbitrate
When to use
Optimised for: not applicable - research paper
Anti-fit: not applicable - research paper
Pros & cons
Pros
Memory-augmented generation with explicit memory tokens.
Cons
Preprint; eval scope limited.
Claims & capabilities
Generative latent memory framework for self-evolving agents with Memory Trigger and Memory Weaver; spontaneously develops planning/procedural/working memory faculties. Headline: surpasses ExpeL and AWM by up to 38.22%; baselines: ExpeL and AWM; primary dataset: 8 benchmarks.
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 (20 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 Generative latent memory vendor
- What it does & what's distinct Weaves generative latent memory for self-evolving agents. vendor
- Claims / benefits Generative latent memory framework for self-evolving agents with Memory Trigger and Memo… vendor
- Modalities text vendor
- Created 2025-09 vendor
- Mindshare 40 cites · 5.6/mo · 367* / 33 forks paper
- Citations 39 cites 39/yr paper
- Founders / pedigree no data — searched arxiv.org/abs/2509.24704; author affiliations not found in preview vendor
- Memory volume / scale no scale claim — research vendor
- Performance +38.22% vs ExpeL/AWM (8 benchmarks, max gain) vendor
- Reproducibility Original team + downstream attention: 33 forks of https://github.com/KANABOON1/MemGen; p… repo
- Code/weights release Code public + weights closed — https://github.com/KANABOON1/MemGen · 367* / 33 forks · l… repo
- Versioning snapshots vendor
- Schema evolution schema locked vendor
- Contradiction handling LLM resolves and overwrites vendor
- Pros Memory-augmented generation with explicit memory tokens. vendor
- Cons Preprint; eval scope limited. vendor
- Project & sources arxiv 2509.24704 vendor
- citation-trajectory [{"year":2025,"cum":1,"infl":0}] paper
- taxonomy-three-layer-rollup agent-infra repo