Lida Chen (陈立达) is a second year Ph.D candidate at Fudan University in the School of Computer Science. His interested research topics are mostly around long context models, including (but are not limited to):

1. Long-Context LLMs: Aiming to enable LLMs to efficiently and effectively process extremely long inputs, reducing the deployment cost of 1M-context production-level transformers to be as affordable as 4K models. This involves integrating techniques such as long-term memory mechanisms, sparse attention, and KV cache compression, among others, into LLMs.
2. Video Language Models: Improving long video understanding capabilities of multi-modal large language models (MLLMs). This involves building datasets for long video understanding, developing vision language connectors for lossless compression of video features, and creating efficient model frameworks for processing long videos.