Profiles2023-2026
Junkang Wu
University of Science and Technology of China
8Papers
1Last author
9Unique co-authors
6First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers8
- Quantile Advantage Estimation: Stabilizing RLVR for LLM ReasoningAuthor 1 of 6ICLR 2026 · 11 citations
- RePO: Understanding Preference Learning Through ReLU-Based OptimizationAuthor 1 of 8NeurIPS 2025 · 4 citations
- Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference OptimizationAuthor 1 of 9ICLR 2025
- β-DPO: Direct Preference Optimization with Dynamic βAuthor 1 of 8NeurIPS 2024 · 114 citations
- Lower-Left Partial AUC: An Effective and Efficient Optimization Metric for RecommendationAuthor 6 of 6WWW 2024 · 13 citations
- BSL: Understanding and Improving Softmax Loss for RecommendationAuthor 1 of 6ICDE 2024 · 11 citations
- Direct Multi-Turn Preference Optimization for Language AgentsAuthor 3 of 5EMNLP 2024 · 1 citation
- Understanding Contrastive Learning via Distributionally Robust OptimizationAuthor 1 of 6NeurIPS 2023 · 55 citations
