Pengfei He

prof_pic.jpg

428 S Shaw Ln Rm 3308

East Lansing, MI, 48824

I am a PhD student major in Computer Science and Engineering and minor in Probability and Statistics, at Michigan State University.

My research interests are robustness and safety of machine learning models, optimization and machine learning foundations. Currently, I am interested in evaluate the robustness of self-/un-supervised learning models, including diffusion models, large language models and text-image foundation models. Looking forward to communicating with people from different fields!

Stay Hungry. Stay Foolish!

news

Nov 8, 2024 Our work Stealthy Backdoor Attack via Confidence-driven Sampling is accepted to TMLR!
Oct 21, 2024 We preprint a new work A Theoretical Understanding of Chain-of-Thought: Coherent Reasoning and Error-Aware Demonstration!
Oct 14, 2024 We preprint a new work Towards the Effect of Examples on In-Context Learning: A Theoretical Case Study!
Oct 10, 2024 One paper(Towards the Effect of Examples on In-Context Learning: A Theoretical Case Study) is accepted to M3L and SFLLM NeruIPS 2024!
Oct 1, 2024 I will serve as the reviewer for ICLR 2025 and AISTAT 2025.
Sep 20, 2024 Two papers( Towards Understanding Jailbreak Attacks in LLMs: A Representation Space Analysis; On the generalization of training-based chatgpt detection methods) accepted to EMNLP 2024!
Jun 3, 2024 I start a new position as Research Intern at Alibaba Group(US) in Bellevue, WA.
May 16, 2024 We have two papers( Exploring Memorization in Fine-tuned Language Models; The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG)) accepted to ACL2024!
Feb 6, 2024 We preprint a paper: Data Poisoning for In-context Learning
Feb 6, 2024 We preprint a paper: Superiority of Multi-Head Attention in In-Context Linear Regression
Feb 5, 2024 We release our survey paper about copyright: Copyright Protection in Generative AI: A Technical Perspective
Jan 16, 2024 Our paper: Sharpness-aware Data Poisoning Attack is accepted as Spotlight (5%) by ICLR2024!
Oct 11, 2023 We preprint a paper: Exploring Memorization in Fine-tuned Language Models.
Oct 10, 2023 We preprint a paper: On the Generalization of Training-based ChatGPT Detection Methods.
Oct 10, 2023 We preprint a paper: FT-Shield: A Watermark Against Unauthorized Fine-tuning in Text-to-Image Diffusion Models.
Oct 9, 2023 We preprint a paper: Confidence-driven Sampling for Backdoor Attacks.
Sep 8, 2023 Our paper Analyzing Illegal Psychostimulant Trafficking Networks Using Noisy and Sparse Data is on IISE Transactions now.
Jul 22, 2023 I will serve as an external reviewer for ICDM 2023.
Jul 13, 2023 I will serve as the PC member of AAAI’24.
May 25, 2023 We preprint a paper: DiffusionShield: A Watermark for Copyright Protection against Generative Diffusion Models.
May 24, 2023 We preprint a paper: Sharpness-aware Data Poisoning Attack.
Apr 24, 2023 Our paper Probabilistic Categorical Adversarial Attack & Adversarial Training is accepted to ICML2023.
Apr 20, 2023 Our paper Large sample spectral analysis of graph-based multi-manifold clustering is accepted to Journal of Machine Learning Research.
Dec 29, 2022 I will serve as the PC member of KDD’23.
Sep 28, 2022 We preprint a paper: Probabilistic Categorical Adversarial Attack & Adversarial Training.
Aug 15, 2022 We hold a lecture-style tutorial about Adversarial Robustness and Poisoning Attacks in the KDD 2022.
Aug 10, 2022 I will serve as the PC member of AAAI’23.
Aug 1, 2022 Our paper PROPN: Personalized Probabilistic Strategic Parameter Optimization in Recommendations got accepted to CIKM’22.
Jul 14, 2021 We preprint a paper: Large sample spectral analysis of graph-based multi-manifold clustering.