arXiv subjects

Jihyung Lee

Publications and source records attributed to Jihyung Lee.

4 recordsLinked to original sources

arXiv2026

EXPO-SQL: Execution-based Clause-level Policy Optimization for Text-to-SQL

Text-to-SQL enables users to query databases using natural language by generating executable SQL queries. Recent methods have increasingly adopted Large Language Models based reinforcement learning (RL) to leverage execution feedback for training. However, existing RL methods assign uniform query-level rewards to all clauses in a SQL query, treating correct and incorrect clauses equally. This coarse-grained reward design leads to insufficient learning signals for correct SQL generation. To address this issue, we propose EXPO-SQL (EXecution-based clause-level Policy Optimization for Text-to-SQL) which provides fine-grained supervision through clause-level rewards. To assign clause-level rewards, our method identifies erroneous clauses by analyzing execution results, including error messages and clause-wise incremental execution. Experiments on widely-used Text-to-SQL benchmarks demonstrate that EXPO-SQL significantly outperforms existing supervised fine-tuning, prompting, and RL-based methods through fine-grained clause-level learning. Our code is available at https://github. com/jhn25/EXPO-SQL.

cs.CL↗

arXiv2025

KoBLEX: Open Legal Question Answering with Multi-hop Reasoning

Large Language Models (LLM) have achieved remarkable performances in general domains and are now extending into the expert domain of law. Several benchmarks have been proposed to evaluate LLMs' legal capabilities. However, these benchmarks fail to evaluate open-ended and provision-grounded Question Answering (QA). To address this, we introduce a Korean Benchmark for Legal EXplainable QA (KoBLEX), designed to evaluate provision-grounded, multi-hop legal reasoning. KoBLEX includes 226 scenario-based QA instances and their supporting provisions, created using a hybrid LLM-human expert pipeline. We also propose a method called Parametric provision-guided Selection Retrieval (ParSeR), which uses LLM-generated parametric provisions to guide legally grounded and reliable answers. ParSeR facilitates multi-hop reasoning on complex legal questions by generating parametric provisions and employing a three-stage sequential retrieval process. Furthermore, to better evaluate the legal fidelity of the generated answers, we propose Legal Fidelity Evaluation (LF-Eval). LF-Eval is an automatic metric that jointly considers the question, answer, and supporting provisions and shows a high correlation with human judgments. Experimental results show that ParSeR consistently outperforms strong baselines, achieving the best results across multiple LLMs. Notably, compared to standard retrieval with GPT-4o, ParSeR achieves +37.91 higher F1 and +30.81 higher LF-Eval. Further analyses reveal that ParSeR efficiently delivers consistent performance across reasoning depths, with ablations confirming the effectiveness of ParSeR.

cs.CL↗

arXiv2025

DCG-SQL: Enhancing In-Context Learning for Text-to-SQL with Deep Contextual Schema Link Graph

Text-to-SQL, which translates a natural language question into an SQL query, has advanced with in-context learning of Large Language Models (LLMs). However, existing methods show little improvement in performance compared to randomly chosen demonstrations, and significant performance drops when smaller LLMs (e.g., Llama 3.1-8B) are used. This indicates that these methods heavily rely on the intrinsic capabilities of hyper-scaled LLMs, rather than effectively retrieving useful demonstrations. In this paper, we propose a novel approach for effectively retrieving demonstrations and generating SQL queries. We construct a Deep Contextual Schema Link Graph, which contains key information and semantic relationship between a question and its database schema items. This graph-based structure enables effective representation of Text-to-SQL samples and retrieval of useful demonstrations for in-context learning. Experimental results on the Spider benchmark demonstrate the effectiveness of our approach, showing consistent improvements in SQL generation performance and efficiency across both hyper-scaled LLMs and small LLMs. The code is available at https://github.com/jjklle/DCG-SQL}{https://github.com/jjklle/DCG-SQL.

cs.AI↗

arXiv2017

QuickTalk: An Association-Free Communication Method for IoT Devices in Proximity

IoT devices are in general considered to be straightforward to use. However, we find that there are a number of situations where the usability becomes poor. The situations include but not limited to the followings: 1) when initializing an IoT device, 2) when trying to control an IoT device which is initialized and registered by another person, and 3) when trying to control an IoT device out of many of the same type. We tackle these situations by proposing a new association-free communication method, QuickTalk. QuickTalk lets a user device such as a smartphone pinpoint and activate an IoT device with the help of an IR transmitter and communicate with the pinpointed IoT device through the broadcast channel of WiFi. By the nature of its association-free communication, QuickTalk allows a user device to immediately give a command to a specific IoT device in proximity even when the IoT device is uninitialized, unregistered to the control interface of the user, or registered but being physically confused with others. Our experiments of QuickTalk implemented on Raspberry Pi 2 devices show that the end-to-end delay of QuickTalk is upper bounded by 2.5 seconds and its median is only about 0.74 seconds. We further confirm that even when an IoT device has ongoing data sessions, QuickTalk can still establish a reliable communication channel to the IoT device with little impact to the ongoing sessions.

cs.NI↗