talks

I gave themed talks on our research works. Here are some examples.

• Your Next-Token Prediction and Transformers Are Biased for Long-Context Modeling
  Invited Talk at the ASAP seminar (slides and recording), Virtual (June 2025)
  Invited Talk at Xiaomi AI Lab Foundation Model team, Virtual (June 2025)

• Contextual Self-supervised Learning: A Lesson from LLMs
  Invited Talk at the Self-Supervised Learning Workshop, Flatiron Institute (Simons Foundation) (recording) (Apr 2025)

• Self-learning Principles of Large-scale Foundation Models
  Invited Talk at the CSE 600 Seminar, Stony Brook University (April 2025)
  Invited Talk at the CILVR seminar at CDS, New York University (Feb 2025)
  Invited Talk at Boston University (Jan 2025)
  Invited Talk at MINDS Jr Seminar, John Hopkins University, US (Dec 2024)
  

• Towards Test-time Self-supervised Learning
  Guest Lecture (slides), CSCI 3370: Deep Learning, Boston College, US (Nov 2024)

• Reimagining Self-supervised Learning with Context
  Invited Talk at MIT ML Tea Seminar, US (Oct 2024)
  Invited Talk at Princeton University, US (Aug 2024)

• Building Safe Foundation Models from Principled Understanding
  Invited Talk at New York University Tandon, US (Sep 2024)

• Non-negative Contrastive Learning
  Invited Talk at Cohere AI, Virtual (Jun 2024)
  Invited Talk at ML Tea, MIT, US (Apr 2024)

• Self-supervised Learning of Identifiable Features
  Invited Talk at TU Munich, Germany (May 2024)

• Understanding and Applying Self-supervised Learning via Graph
  Invited Talk at Deep Potential, China (2023)
  Invited Talk at KAIST, Virtual (2022)

• Towards Truly Unlearnable Examples for Data Privacy
  Invited Talk at Chinese Academy of Sciences, China (2022)

• Reparameterized Sampling for GANs
  Invited Talk at Beijing Academy of Artificial Intelligence (BAAI), China (2021)
  Plenary Talk at ECML-PKDD 2021, Virtual (2021)