5 papers
Humanity's Last Code Exam: Can Advanced LLMs Conquer Human's Hardest Code Competition?
Xiangyang Li, Xiaopeng Li, Kuicai Dong +7
Code generation is a core capability of large language models (LLMs), yet mainstream benchmarks (e.g., APPs and LiveCodeBench) contain questions with medium-level difficulty and po…
MAPO: Mixed Advantage Policy Optimization
Wenke Huang, Quan Zhang, Yiyang Fang +11
Recent advances in reinforcement learning for foundation models, such as Group Relative Policy Optimization (GRPO), have significantly improved the performance of foundation models…
AALC: Large Language Model Efficient Reasoning via Adaptive Accuracy-Length Control
Ruosen Li, Ziming Luo, Quan Zhang +4
Large reasoning models (LRMs) achieve impressive reasoning capabilities by generating lengthy chain-of-thoughts, but this "overthinking" incurs high latency and cost without commen…
Beyond Invisibility: Learning Robust Visible Watermarks for Stronger Copyright Protection
Tianci Liu, Tong Yang, Quan Zhang +1
As AI advances, copyrighted content faces growing risk of unauthorized use, whether through model training or direct misuse. Building upon invisible adversarial perturbation, recen…
Elastic Representation: Mitigating Spurious Correlations for Group Robustness
Tao Wen, Zihan Wang, Quan Zhang +1
Deep learning models can suffer from severe performance degradation when relying on spurious correlations between input features and labels, making the models perform well on train…