Bio

I am an AI researcher at Meta Superintelligence Labs (MSL). I received my B.S., M.S., and Ph.D. from UC Santa Cruz, where I worked with Prof. Yi Zhang on Natrual Language Processing. During my PhD studies, I interned at Google DeepMind, Amazon Alexa, Google X, and multiple AI start-ups.

My research aims to empower AI systems with advanced reasoning capabilities. I work at the intersection of the following topics:
  • Complex Planning and Reasoning: My research aims to solve real-world problems, which often require multi-step reasoning and actions.
  • Multi-agent Collaboration: I work on MAS (multi-agent systems), so my study also includes how agents coordinate, communicate, and share information.
  • Information Retrieval: I conduct research on, but not limited to, RAG with LLM; broadly speaking, I care about Retrieval-Augmented AI.

News

Experience

• Meta MSL: Research Scientist, 2026
• Google DeepMind: Student Researcher, 2025
• Mineral.ai: AI Resident, 2024
• Amazon Alexa: Applied Scientist Intern, 2023
• Mineral.ai, an Alphabet company, previous Google[X] project: AI Resident, 2023

Teaching and Mentoring

Research Mentor:
  I am dedicated to mentoring students, ranging from high schoolers to Grad students, supporting them in pursuing their career goals in either academic research or engineering in AI.

Instructor:
  Fall 2025 UC Santa Cruz: NLP 255 – Topics in Applied Natural Language Processing.

Services

Reviewer:
  2026 - AAAI, ECIR, ICLR, ICML, SIGIR, ACL, NeurIPS
  2025 - ECIR, ICLR, ICML, SIGIR, ACL, NeurIPS, Knowledge-Based Systems
  2024 - TheWebConf (WWW), CVPR, ICML, NeurIPS
  2023 - NeurIPS

Event Organizer:
  Co-organized the tools-meet-llm tutorial at ACM SIGIR 2024.
  Co-chair and organizer of the 1st AI Student Research Symposium at UC Santa Cruz.

Selected Publications

^*Equal contribution, ^†Co-advising.

Full publication list can be found on Google Scholar page.

Don’t Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-head Decoding

Yunkai Zhang, Qiang Zhang, Feng Lin, Ruizhong Qiu, Hanchao Yu, Jiayi Liu, Yinglong Xia, Zhuoran Yu, Zeyu Zheng, Diji Yang

WWW 2026 Oral [Paper] [Code]

---

GenIR: Generative Visual Feedback for Mental Image Retrieval

Diji Yang^*, Minghao Liu^*, Chung-Hsiang Lo, Yi Zhang, James Davis

NeurIPS 2025 [Paper] [Code]

---

Worse than Zero-shot? A Fact-Checking Dataset for Evaluating the Robustness of RAG Against Misleading Retrievals

Linda Zeng, Rithwik Gupta, Divij Motwani, Yi Zhang^†, Diji Yang^†

NeurIPS 2025 [Paper] [Dataset]

---

Knowing You Don't Know: Learning When to Continue Search in Multi-round RAG through Self-Practicing

Diji Yang^*, Linda Zeng^*, Jinmeng Rao, Yi Zhang

SIGIR 2025 [Paper] [Code]

---

Right this way: Can VLMs Guide Us to See More to Answer Questions?

Li Liu^*, Diji Yang^*, Sijia Zhong, Kalyana Suma Sree Tholeti, Lei Ding, Yi Zhang, Leilani H. Gilpin

NeurIPS 2024 [Paper] [Code] [Dataset]

---

IM-RAG: Multi-Round Retrieval-Augmented Generation Through Learning Inner Monologues

Diji Yang, Jinmeng Rao, Kezhen Chen, Xiaoyuan Guo, Yawen Zhang, Jie Yang, Yi Zhang

SIGIR 2024 [Paper]

---

Tackling vision language tasks through learning inner monologues

Diji Yang, Jinmeng Rao, Kezhen Chen, Xiaoyuan Guo, Yawen Zhang, Jie Yang, Yi Zhang

AAAI 2024 [Paper]

---

Get Connected

I am always interested in working with new collaborators from all backgrounds on research ideas and/or impactful projects related to my areas of interest. That being said, I commit the sufficient time to each project I take on, so my availability for new projects may depends on current commitments.

If you’re interested in collaborating, feel free to send me an email to find a time to chat!