Scholar
Ru Peng
Google Scholar ID: 3udA8hkAAAAJ
Zhejiang University & Qwen Team, Alibaba Group
Citations & Impact
Citations
5,120
H-index
9
i10-index
9
Publications
20
Co-authors
12
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'LLM-Enhanced Query Generation and Retrieval Preservation for Task-Oriented Dialogue' accepted at Findings of ACL 2025
- Paper 'DataMan: Data Manager for Pre-training Large Language Models' accepted at ICLR 2025
- Paper 'Inference-Time Decontamination: Reusing Leaked Benchmarks for Large Language Model Evaluation' accepted at Findings of EMNLP 2024
- Two papers 'Predicting Rewards Alongside Tokens' and 'Embedding and Gradient Say Wrong' accepted at EMNLP 2024
- Paper 'DORY: Deliberative Prompt Recovery for LLM' accepted at Findings of ACL 2024
- Paper 'Energy-based Automated Model Evaluation' accepted at ICLR 2024
- Paper 'CAME: Contrastive Automated Model Evaluation' accepted at ICCV 2023
- Paper 'Distill The Image to Nowhere' accepted at EMNLP 2022 (Oral)
- Paper 'HybridVocab: Towards Multi-Modal Machine Translation via Multi-Aspect Alignment' accepted at ICMR 2022 (Oral)
- Contributed to the release of Qwen series foundation models (Qwen1.5, Qwen2, Qwen2.5, Qwen3) and technical reports
- Involved in the Dotamath project for mathematical reasoning
Background
- 4th-year PhD student at the College of Computer Science and Technology, Zhejiang University
- Research interests span Large Language Models (current focus), Machine Learning, NLP, and Multimodal AI, aiming to build AGI to transform human life
- Focuses on pre-training data management and data synthesis for LLMs
- Develops contrastive and energy-based unsupervised model evaluation methods
- Works on multimodal, sign language, and text-only machine translation
