Generative Agents
https://arxiv.org/abs/2304.03442
Park et al. — landmark agent-simulation paper. Reflection + memory stream + retrieval enable believable agent behavior.
Research code · Verified 2026-10-23 · fresh as of 2026-10-23 · 28 cited facts, 1 with a third-party source · 1 recorded change · Suggest a correction
At a glance
- Type
- Interactive simulacra of human behavior vendor fix
- Tier
- T3
- Latest release
- no releases fix
- License
- Apache-2.0 repo fix
- Pricing
- not applicable — research paper fix
- Funding
- not applicable — not commercial fix
- Freshness
- 2026-10-23 fix
- Layer
- agent-infra repo fix
- Licensing
- open-source repo fix
- Developer activity
- OpenRank 0.43 (12-mo avg 0.85) · 1 participants — 2026-09 third-party fix
Memory taxonomy
- storage
- vector
- retrieval
- similarity
- persistence
- long-term
- update
- append-only
- unit
- episode
- governance
- inspectable
- conflict
- llm-arbitrate
When to use
Optimised for: not applicable - research paper
Anti-fit: not applicable - research paper
Pros & cons
Pros
Stanford Park et al. — defined the LLM-simulation-of-society field; cited by every subsequent agent-memory paper.
Cons
Simulation-only; not a production memory system; eval is qualitative.
Claims & capabilities
Park et al. simulacra of human behavior in a sandbox town with 25 agents emergent social dynamics. Headline: case-study of believable behavior (Valentine's Day party emergence) — no quantitative headline in abstract; baseline: ablations of memory/reflection components; primary dataset: custom Smallville sandbox (25 agents in The Sims-style environment).
Technical surface
- API surface
- not applicable — research paper
- Backend storage
- not applicable — research paper
- Deployment
- not applicable — research paper
- 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.
- MemoryBank T3
Enhances LLMs with long-term memory. Early influential paper.
- Reflexion T3
Language agents with verbal reinforcement learning.
- DreamerV3 T3
Hafner et al. General world-model agent. Symlog observation, KL balance, RMSNorm, Block GRU.
Sources (27 cited facts, 1 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is Interactive simulacra of human behavior vendor
- What it does & what's distinct Park et al. — landmark agent-simulation paper. Reflection + memory stream + retrieval en… vendor
- Claims / benefits Park et al. simulacra of human behavior in a sandbox town with 25 agents emergent social… vendor
- Modalities text (natural language simulation) vendor
- Created 2023-07 vendor
- License Apache-2.0 repo
- GitHub 21.2k★ repo
- Mindshare 3905 cites · 105.7/mo · 21257* / 2990 forks · HN front page #35517649 + #37073938; UIST'… paper
- Citations 3.9k cites 1298/yr paper
- Founders / pedigree Stanford HCI + Stanford NLP; Joon Sung Park, Joseph C. O'Brien, Carrie J. Cai, Meredith … vendor
- Memory volume / scale 25 agents in Smallville sandbox vendor
- Performance 25 agents in Smallville sandbox emergent social dynamics (Valentine's Day party self-org… vendor
- Reproducibility Independently reproduced: official repo + a16z-infra/ai-town + Project Sid (arXiv:2411.0… repo
- Code/weights release Code public, weights closed (uses GPT API; no model weights) — https://github.com/joonsp… repo
- Versioning immutable append-log vendor
- Tombstoning append-only no delete vendor
- Schema evolution schema with migration vendor
- Contradiction handling always append (no resolution) vendor
- Forgetting policy none (permanent) vendor
- Pros Stanford Park et al. — defined the LLM-simulation-of-society field; cited by every subse… vendor
- Cons Simulation-only; not a production memory system; eval is qualitative. vendor
- Project & sources arxiv 2304.03442 GitHub joonspk-research vendor
- commit-trajectory [{"ym":"2023-07","cum":19},{"ym":"2023-08","cum":31}] repo
- citation-trajectory [{"year":2023,"cum":10,"infl":3},{"year":2024,"cum":18,"infl":6},{"year":2025,"cum":27,"… paper
- licensing-bucket open-source repo
- taxonomy-three-layer-rollup agent-infra repo
- developer-activity-opendigger {"activity":2.02,"as_of":"2026-09","openrank_12mo":0.85,"openrank_current":0.43,"partici… third-party
What changed
Changes to this record's lifecycle, licence, release, pricing and hosting facts, from the catalog's git history. Subscribe (Atom)
- 2026-07-02
- Lifecycle first recorded as “research-complete”
Related systems
Referenced by (7)
- Fincon cites — S2 isInfluential citation
- G-Memory cites — S2 isInfluential citation
- LoCoMo cites — S2 isInfluential citation
- Lyfe Agents cites — S2 isInfluential citation
- MemTree cites — S2 isInfluential citation
- Retroformer cites — S2 isInfluential citation
- S³ cites — S2 isInfluential citation