CV

Last updated Oct 2026. PDF available via the button on the right.

Contact Information

Full Name Xin Su
Email xinsu626@gmail.com

Education

  • 2020 - 2024
    Ph.D. in Information
    University of Arizona, Tucson, AZ

Experience

  • 11/2025 - Present
    Lead AI Researcher
    Thoughtworks
  • 01/2025 - 11/2025
    AI Research Scientist
    Intel Labs
  • 05/2023 - 12/2024
    AI Research Intern
    Intel Labs
  • 05/2022- 08/2022
    AI Research Intern
    Intel Labs, USA
  • 2020 - 2024
    Graduate Research Associate
    Computational Language Understanding Lab, University of Arizona

Recent Publications (2026)

  • Fangyuan Yu, Xin Su, Amir Abdullah. Dynamic Latent Routing. NeurIPS 2026.
  • Phillip Howard, Xin Su, Kathleen C. Fraser. Cross-Cultural Value Attribution in Large Vision-Language Models. Findings of EMNLP 2026.
  • Prafulla Kumar Choubey*, Xin Su*, Man Luo*, et al. Scaling Knowledge Graph Construction through Synthetic Data Generation and Distillation. ICLR 2026. (*Equal contribution)
  • Phillip Howard, Xin Su. Culture-Dependent Refusal Disparities in Large Vision-Language Models. VLM4RWD Workshop at NeurIPS 2026.
  • Fateme Hashemi Chaleshtori, Fangyuan Yu, Xin Su, Amir Abdullah, Phillip Howard. TRAP: Auditing Language Models for Conjunctive, Non-Adjacent Backdoor Triggers. NeuralArtifacts Workshop at NeurIPS 2026.
  • Phillip Howard, Xin Su, Allen Roush, Manikandan Ravikiran, Amir Abdullah. Refusal-Gated Decoding: Preserving Refusal Behavior Under High-Temperature Sampling. AdvML-Frontiers × CoTMA Workshop at COLM 2026.
  • Phillip Howard, Xin Su, Kathleen C. Fraser. Cultural Counterfactuals: Evaluating Cultural Biases in Large Vision-Language Models with Counterfactual Examples. SAFE Workshop at CVPR 2026; TrustNLP Workshop at ACL 2026.
  • Ankit Pratap Singh, Xin Su, Phillip Howard. Synthetic Contrastive Reasoning for Multi-Table Q&A. SURGeLLM Workshop at ACL 2026.

Academic Service

  • Area Chair: ACL Rolling Review (ARR), 2026 - Present
  • Reviewer / Program Committee: NeurIPS (2023, 2025), ICML (2024 - 2026), ICLR (2024 - 2027), AAAI (2025, 2027), AISTATS 2026, ACL 2023, EMNLP (2022, 2023), ARR (2023 - Present), TMLR (2026 - Present)