Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.11333 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914386808930304 |
|---|---|
| author | Zhang, Haoting Lin, Yunduan He, Jinghai Jiang, Denglin Zuo-Jun Shen Zheng, Zeyu |
| author_facet | Zhang, Haoting Lin, Yunduan He, Jinghai Jiang, Denglin Zuo-Jun Shen Zheng, Zeyu |
| contents | Short-video platforms are closed-loop, human-in-the-loop ecosystems where platform policy, creator incentives, and user behavior co-evolve. This feedback structure makes counterfactual policy evaluation difficult in production, especially for long-horizon and distributional outcomes. The challenge is amplified as platforms deploy AI tools that change what content enters the system, how agents adapt, and how the platform operates. We propose a large language model (LLM)-augmented digital twin for short-video platforms, with a modular four-twin architecture (User, Content, Interaction, Platform) and an event-driven execution layer that supports reproducible experimentation. Platform policies are implemented as pluggable components within the Platform Twin, and LLMs are integrated as optional, schema-constrained decision services (e.g., persona generation, content captioning, campaign planning, trend prediction) that are routed through a unified optimizer. This design enables scalable simulations that preserve closed-loop dynamics while allowing selective LLM adoption, enabling the study of platform policies, including AI-enabled policies, under realistic feedback and constraints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_11333 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | LLM-Augmented Digital Twin for Policy Evaluation in Short-Video Platforms Zhang, Haoting Lin, Yunduan He, Jinghai Jiang, Denglin Zuo-Jun Shen Zheng, Zeyu Artificial Intelligence Short-video platforms are closed-loop, human-in-the-loop ecosystems where platform policy, creator incentives, and user behavior co-evolve. This feedback structure makes counterfactual policy evaluation difficult in production, especially for long-horizon and distributional outcomes. The challenge is amplified as platforms deploy AI tools that change what content enters the system, how agents adapt, and how the platform operates. We propose a large language model (LLM)-augmented digital twin for short-video platforms, with a modular four-twin architecture (User, Content, Interaction, Platform) and an event-driven execution layer that supports reproducible experimentation. Platform policies are implemented as pluggable components within the Platform Twin, and LLMs are integrated as optional, schema-constrained decision services (e.g., persona generation, content captioning, campaign planning, trend prediction) that are routed through a unified optimizer. This design enables scalable simulations that preserve closed-loop dynamics while allowing selective LLM adoption, enabling the study of platform policies, including AI-enabled policies, under realistic feedback and constraints. |
| title | LLM-Augmented Digital Twin for Policy Evaluation in Short-Video Platforms |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.11333 |