EPISODE · Jun 17, 2026 · 6 MIN
AI计算广告论文播报｜6月16日：多轮改稿RL上线搜索广告与Agent状态校验范式
from 周六9点半
本期聚焦两个核心问题：广告文案生成如何从"一次写完"进化到"多轮自我修订"？LLM Agent 的输出如何不污染系统状态？同时我们也观察到多篇论文在集体质疑检索系统的"容量幻觉"。本期重点* Interactor: Agentic RL oriented Iterative Creation for Ad Description Generation in Sponsored Search——百度搜索广告团队将 Agentic RL 用于广告描述的多轮迭代生成，核心亮点是让奖励模型不只打分、还写"批改意见"，模型读着评语改稿，忠实度从 0.73 涨到 0.87。已覆盖 14 万广告主线上部署，广告收入 +0.74%。* Orchestrated Reality: LLM-Driven World Simulation as a Parameterized-Action POMDP——...去小宇宙查看完整单集简介在小宇宙查看该单集文稿
Embed this episode
Ready to play
AI计算广告论文播报｜6月16日：多轮改稿RL上线搜索广告与Agent状态校验范式
0:00
6:54
1×
No transcript for this episode yet
Similar Episodes
No similar episodes found.
Similar Podcasts
No similar podcasts found.
Frequently Asked Questions
How long is this episode of 周六9点半?
This episode is 6 minutes long.
When was this 周六9点半 episode published?
This episode was published on June 17, 2026.
Can I download this 周六9点半 episode?
Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!