Large Language Models (LLMs) have demonstrated impressive ability in generation and reasoning tasks but struggle with handling up-to-date knowledge, leading to inaccuracies or hallucinations. Retrieval-Augmented Generation (RAG) mitigates this by retrieving and incorporating external knowledge into input prompts. In particular, due to LLMs' context window limitations and long-context hallucinations, only the most relevant "chunks" are retrieved. However, current RAG systems face three key challenges: (1) chunks are often retrieved independently without considering their relationships, such as redundancy and ordering; (2) the utility of chunks is non-monotonic, as adding more chunks can degrade quality; and (3) retrieval strategies fail to adapt to the unique characteristics of different queries. To overcome these challenges, we design a cost-constrained retrieval optimization framework for RAG. We adopt a Monte Carlo Tree Search (MCTS) based strategy to find the optimal chunk combination order, which considers the chunks' correlations. In addition, to address the non-monotonicity of chunk utility, instead of treating budget exhaustion as the termination condition, we design a utility computation strategy to identify the optimal chunk combination without necessarily exhausting the budget. Furthermore, we propose a configuration agent that predicts optimal configurations for each query domain, improving our framework's adaptability and efficiency. Experimental results demonstrate up to a 30% improvement over baseline models, highlighting the framework's effectiveness, scalability, and suitability. Our source code has been released at https://github.com/wang0702/CARROT.

In recent years, Large Language Models (LLMs) have transformed computer science education by offering instant code generation, yet these tools often encourage passive completion where students rely on trial-and-error rather than deep algorithmic design. To address this limitation, we propose Design Before Code, a novel interactive framework that enforces a validated visual plan before implementation to bridge the gap between high-level intent and low-level syntax. Our system utilizes a Visual Control Flow Graph analysis powered by Personalized PageRank, which identifies the structurally critical regions of an algorithm by assigning higher weights to complex decision and loop structures. To mitigate dependency, we introduce an interactive agent that gates the code editor based on these structural weights, providing inquiry-based hints instead of full solutions. In a controlled study with 120 students across five diverse algorithmic tasks, Design Before Code achieved an approximately 80% Logic Preservation Rate, exhibiting a significant improvement of 24% compared to the LLM-only baseline. Furthermore, our approach reduced unproductive debugging cycles by 60% while achieving an 86% score on independent transfer tasks, demonstrating that structural scaffolding fosters superior long-term conceptual retention compared to traditional methods.

Deploying database services on cloud systems has gained increasing popularity and has become a common practice in the industry. However, the complicated cloud environments make performance issues inevitable, which could violate the service level guarantee if not addressed in a timely manner. Among the various problems, anomalies in SQL queries are the most commonly reported sources that cause performance issues in database applications. These anomalous queries can be divided into High-impact SQLs (H-SQLs) and Root Cause SQLs (R-SQLs), representing the related SQLs that are correlated with the anomalies and the ones that are the root causes of the performance issue, respectively. In the presence of a large number of queries, to pinpoint the R-SQLs is far more difficult than to identify the H-SQLs. To address this challenge, we aim at automatically pinpointing the R-SQLs to resolve performance issues in cloud databases. This paper introduces PinSQL, an autonomous diagnosing system for Alibaba Cloud, which has four modules that are executed sequentially, including data collection and pre-processing, anomaly detection, root cause analysis, and repairing actions. First, the related performance metrics and query logs from monitored cloud database instances are collected and aggregated as the data sources. Then, based on these inputs, efficient anomaly detection is conducted in real-time. Upon the detection of an anomaly, the root cause SQLs are pinpointed through tracking the propagation chain of the involved SQLs. Finally, repairing actions are suggested and then executed on R-SQLs to address the anomalies. Extensive experiments on an Alibaba production system show that PinSQL can achieve an 80% accuracy for pinpointing the top-1 R-SQLs and successfully resolve the database performance issues resultantly.