Zhaowei (James) Wang

Research Highlights

Multimodal Long Context

1. EMNLP

   Can Retrieval Heads See Images? Multimodal Retrieval Heads in Long-Context Vision-Language Models (MMRetHeads)

   In The 2026 Conference on Empirical Methods in Natural Language Processing

   arXiv
2. NeurIPS

   Training Long-Context Vision-Language Models Effectively with Generalization Beyond 128K Context (MMProLong)

   In Advances in Neural Information Processing Systems

   arXiv
3. NeurIPS

   MMLongBench: Benchmarking Long-Context Vision-Language Models Effectively and Thoroughly (*spotlight)

   In The 39th (2025) Annual Conference on Neural Information Processing Systems

   arXiv Code Website

Long-Term Memory

1. ICLR

   Memory-T1: Reinforcement Learning for Temporal Reasoning in Multi-session Agents

   arXiv
2. arXiv

   MEMLENS: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models

   arXiv
3. arXiv

   Rethinking Memory in AI: Taxonomy, Operations, Topics, and Future Directions

   arXiv Code