# Affordable Generative Agents

*Source* · `src-yu-2024-affordable-generative-agents` · status: **n/a**

| field | value |
|---|---|
| kind | Preprint |
| authors | Yangbin Yu, Qin Zhang, Junyou Li, Qiang Fu, Deheng Ye |
| venue | arXiv |
| date | 2024-02 |
| url | https://arxiv.org/abs/2402.02053 |
| arxiv | 2402.02053 |
| accessed | 2026-10-24 |

## Notes

Lifestyle Policy replays stored plan-action sequences when a new plan matches by embedding cosine >= 0.97 and conditions hold; Social Memory compresses about 2,000 prompt tokens (30-45 events) to about 100. Tokens: 31.1% of baseline (3 agents, two game days), 42.7% (25 agents), 3.4% (VirtualHome, one agent-day); task success differences not significant.

## Backlinks (computed)

**cited by:** [Generative agents (memory stream, reflection, planning)](../technique/generative-agents.md)

## Citation

Yangbin Yu, Qin Zhang, Junyou Li, Qiang Fu, Deheng Ye (2024-02). *Affordable Generative Agents*. arXiv https://arxiv.org/abs/2402.02053

---
Provenance: created in run `2026-10-24-techniques-llm-agents`.