[人人能懂] 从成本路由到自适应学习

00:00:33 你的AI管家，如何做到既聪明又省钱？

00:05:20 如何让机器像老师傅一样，越老越“精”

00:10:10 AI的新难题：既要会推理，又要会打分

00:15:03 给大象减肥：AI模型的“瘦身”新思路

00:19:15 让“写作高手”变身“超级图书管理员”

本期介绍的五篇论文：

[LG] Adaptive LLM Routing under Budget Constraints  

[Fujitsu Research]  

https://arxiv.org/abs/2508.21141  

---

[LG] CALM: A Framework for Continuous, Adaptive, and LLM-Mediated Anomaly Detection in Time-Series Streams  

[Dataflow ML Team]  

https://arxiv.org/abs/2508.21273  

---

[CL] Reasoning-Intensive Regression  

[MIT]  

https://arxiv.org/abs/2508.21762  

---

[LG] QR-LoRA: QR-Based Low-Rank Adaptation for Efficient Fine-Tuning of Large Language Models  

[University of Pennsylvania]  

https://arxiv.org/abs/2508.21810  

---

[CL] Efficient Code Embeddings from Code Generation Models  

[MIT & Jina AI GmbH]  

https://arxiv.org/abs/2508.21290