I gave a talk at Qualcomm AI Security Lecture Series, March 8, 2024. The talk title is "Invisible Foes: Crafting and Cracking AI in the Shadows of Language -- Poison Finetuning Data and Jailbreak Prompts for LLMs".

I gave a seminar talk at Values-Centered Artificial Intelligence (VCAI) seminar series, College Park, MD, Feb 1, 2024. The talk title is "Algorithmic Fairness in an Ever-Changing World".

Our Mementos benchmark on testing sequential reasoning capabilities of multimodal large language models on image sequences is out. Find the arXiv, github code and data, visualization and leaderboard links at the project page: https://mementos-bench.github.io/. Jan., 2024.

Our WAVES benchmark on stress-testing image watermarks is out. Find the arXiv, github code, Hugging Face data, visualization and leaderboard links at the project page: https://wavesbench.github.io/. Jan., 2024.

10 papers accepted to the main conference of ICLR 2024, 2 as spotlights 8 as posters. For more details, click on the Post on Social Media.

Chair and organizer of NSF-Amazon Fairness in AI Principle Investigator Meeting, Jan 9-10, 2024.

4 papers accepted to the main conference of NeurIPS 2023 and 11 papers accepted to the workshops of NeurIPS 2023, 1 as oral, 2 as spotlights, and 8 as posters, Sep-Dec 2023.

Depart Colloquium at University of Maryland, College Park, "Trustworthy Machine Learning in an Ever-Changing World," Sep., 2023.

Panelist at Interactive Learning with Implicit Human Feedback workshop, ICML, Jul., 2023.

3 papers accepted to the main conference and 7 accepted to the workshops at ICML 2023.

Keynote speaker at ROADS to Mega-AI Models Workshop, "Efficient Machine Learning at the Edge," MLSys, Jun., 2023.

Invited talk at the 3rd Workshop of Adversarial Machine Learning on Computer Vision: Art of Robustness, ``Robust Reinforcement Learning in an Ever-Changing World'', CVPR, Jun., 2023.

5 papers accepted to main conference at ICLR 2023 (1 of which as a spotlight oral presentation), see this thread of twitter threads  for an introduction of these works. In addition, 2 papers accepted to ICLR workshops 2023.

Invited talk on "Adaptable Reinforcement Learning in An Ever-Changing World” at the the Reincarnating Reinforcement Learning workshop at ICLR 2023. See a recording of the talk below.

Panelist at Reincarnating RL workshop, ICLR. May., 2023.

Co-organizer of NSF-IEEE workshop: Toward Explainable, reliable, and sustainable machine learning in signal & data science, ``Trustworthy Machine Learning in Complex Environments'', Mar. 2023.

Invited talk at 57th Annual Conference on Information Science and Systems, CISS, ``Efficient Machine Learning at the Edge in Parallel'', Mar., 2023.

Invited talk at 2023 Information Theory and ApplicationsWorkshop, ITA, ``Trustworthy Machine Learning in Complex Environments'', Feb., 2023.

Invited talk at UTSA Matrix AI Seminar, ``Trustworthy Machine Learning in Complex Environments'', Jan., 2023.