3 papers
cs.CL2026
Scaling Data Difficulty: Improving Coding Models via Reinforcement Learning on Fresh and Challenging Problems
Zongqian Li, Tengchao Lv, Shaohan Huang +8
Training next-generation code generation models requires high-quality datasets, yet existing datasets face difficulty imbalance, format inconsistency, and data quality problems. We…
cs.CL2024
MMLU-CF: A Contamination-free Multi-task Language Understanding Benchmark
Qihao Zhao, Yangyu Huang, Tengchao Lv +8
Multiple-choice question (MCQ) datasets like Massive Multitask Language Understanding (MMLU) are widely used to evaluate the commonsense, understanding, and problem-solving abiliti…
cs.CL2024
RedStone: Curating General, Code, Math, and QA Data for Large Language Models
Yaoyao Chang, Lei Cui, Li Dong +13
Pre-training Large Language Models (LLMs) on high-quality, meticulously curated datasets is widely recognized as critical for enhancing their performance and generalization capabil…