Publications

SWE-Perf: Can Language Models Optimize Code Performance on Real-World Repositories?

Nexus: Execution-Grounded Multi-Agent Test Oracle Synthesis

Secure Code Generation via Online Reinforcement Learning with Vulnerability Reward Model

Scaling Code LLM Training and Test-Time Inference via Execution-Free Reward Models

Paper Espresso: From Paper Overload to Research Insight

Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts

Benchmarking LLMs for Unit Test Generation from Real-World Functions

Semantics-Aligned, Curriculum-Driven, and Reasoning-Enhanced Vulnerability Repair Framework

Pro-QuEST: Prompt-chaining Quiz Engine for testing Specialized Technical Product Knowledge

Improving Arabic Dialectness in LLMs with Reinforcement Learning

CodeArena: A collective evaluation platform for LLM code generation

AntiLeakBench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge

Towards Verifiable Text Generation with Generative Agent

Guardreasoner-VL: Safeguarding VLMs via Reinforced Reasoning

Afterburner: Reinforcement Learning Facilitates Self-Improving Code Efficiency Optimization

Unraveling Online Mental Health Through the Lens of Early Maladaptive Schemas: AI-Enabled Content Analysis of Online Mental Health Communities

EffiBench-X: A Multi-Language Benchmark for Measuring Efficiency of LLM-Generated Code

Measuring the Influence of Incorrect Code on Test Generation

Afterburner: Reinforcement Learning Facilitates Self-Improving Code Efficiency Optimization

Guardreasoner-VL: Safeguarding VLMs via Reinforced Reasoning

On Assigning Product and Software Codes to Service Requests with Large Language Models

Position: Current Model Licensing Practices are Dragging Us into a Quagmire of Legal Noncompliance

Curriculum Demonstration Selection for In-Context Learning

From Static to Dynamic: Knowledge Metabolism for Large Language Models

Mercury: A Code Efficiency Benchmark for Code Large Language Models

Counseling Responses for Mental Health Forum Questions with Early Maladaptive Schema Prediction

Chain-of-Thought Improves Text Generation with Citations in Large Language Models

Constituency-Informed and Constituency-Constrained Extractive Question Answering with Heterogeneous Graph Transformer

Identifying Checkworthy Cure Claims on Twitter

Generating Reflective Questions for Engaging Gallery Visitors in ArtMuse

NUS-IDS at CheckThat! 2022: identifying check-worthiness of tweets using CheckthaT5