Shanshan Zhong

Shanshan Zhong

钟珊珊

Ph.D. Student

Carnegie Mellon University

Language Technologies Institute

I am a CS Ph.D. student at Carnegie Mellon University starting 2025 Fall, advised by Prof. Chenyan Xiong. Prior to this, I earned both my master's and bachelor's degrees from Sun Yat-sen University, where I was advised by IEEE Fellow Prof. Liang Lin and Prof. Wushao Wen. I have also had the pleasure of collaborating with Prof. Alan Yuille and Jieneng Chen at CCVL, as well as Prof. Pan Zhou and Shanghua Gao at Sea AI Lab. I also had a great experience contributing to OneFlow. My research interests include:

  LLM Reasoning, Memory, and Agent Optimization
  1. Low-cost long-context benchmarks [MiniLongBench] and LLM routing [RouterEval]
  2. Chain-of-thought search for optimal reasoning paths [Neural CoT Search]
  3. Creative leap-of-thought thinking for multimodal LLMs [CLoT]
  4. Causality-aware creativity evaluation for multimodal LLMs [LoTbench]
  5. Data-driven analysis of agent skills for extending LLM functionality [Agent Skills]
  Web & Information Retrieval
  1. Generative search engine optimization for web content [AutoGEO]
  2. Robust multimodal recommendation via flat local minima [Mirror Gradient]
  3. Variational information bottleneck distillation for multimodal RecSys [DVIB]

News

   Older News

Upcoming Events

Awards

Get Connected

Undergrad & Master's Students: I'm always excited to work with motivated students. If you're interested in exploring a research collaboration or want to chat about potential projects, feel free to email me and we can set up a time to talk!

Fellow Researchers: I love exchanging ideas and finding unexpected connections across research areas. Whether you're working on something related or just want to brainstorm, don't hesitate to reach out — I'm always happy to chat over a virtual coffee ☕

Academic Service

Invited Journal Reviewer: IEEE TKDE, IEEE TMM, IEEE TPAMI

Invited Conference Reviewer: ICLR, ICML, ACL, EMNLP, NeurIPS, WWW, CVPR, ICCV, ECCV, AAAI