My course on differential privacy is public! Check out the course page for: lecture videos 🎥, lecture notes 📝, and suggested readings 📖! gautamkamath.com/CS860-fa2020.h… Be sure to subscribe on Youtube to catch the latest lectures! youtube.com/playlist?list=… #privacy

🔥 VAEBM = VAE + EBM 🔥 VAEBM is a new generative model defined by a symbiotic composition of VAE and EBM. VAEBM improves generative quality of VAEs & explicit EBMs by a large margin & reduces the gap with GANs. arxiv.org/abs/2010.00654 w/ Zhisheng Xiao, Karsten Kreis, Jan Kautz

When I was a PhD student, a professor told me: "There are only three papers worth writing on a given topic: the first paper, the best paper, and the last paper." It's obv a bit tongue-in-cheek, but I've often used the underlying heuristic to aid in choosing research directions.

GNNs are widely believed to work well due to the homophily assumption and recent works design new architectures to overcome such heterophily-related limitations. Is Homophily a Necessity for Graph Neural Networks? the answer is in our new preprint: arxiv.org/abs/2106.06134

Improving Robustness of Graph Neural Networks with Heterophily-Inspired Designs. (arXiv:2106.07767v1 [cs.LG]) ift.tt/3pW3eGi

Thanks Jiliang Tang for sharing this interesting and thought-provoking work. In light of these new findings, we revisit the problem of heterophily for GNNs and discuss the reasons behind the seemingly different takeaways from different works. Check it out: jiongzhu.net/revisiting-het…

Think privacy in ML ➡️ bad accuracy? 😭 Think again! We show that large language models (e.g., RoBERTa and GPT-2) can be privately fine-tuned with minimal loss in utility! 🥳 A super fun Waterloo's Cheriton School of Computer Science Microsoft Research collab! 📝Paper: arxiv.org/abs/2110.06500 🧵Thread ⬇️ 1/8

Tomorrow, I will give a tutorial on "ML from a Bayesian Perspective" at ACML 2021. acml-conf.org/2021/school/sc… I also wrote a summary for the slides, explaining why a Bayesian perspective is essential for machine learners emtiyaz.github.io/papers/MLfromB… Slides: emtiyaz.github.io/papers/Nov9_20…

Arash Vahdat You can also run this script on this years submissions: github.com/evanzd/ICLR202…

Arash Vahdat Ok so I managed to scrape it, you can see my fork: github.com/kashif/ICLR202…

SHAP: Explain Any Machine Learning Model in Python bit.ly/3pQLOwH #DataScience #MachineLearning #AI #DeepLearning

📢 1/ We are delighted to have Nicholas Carlini for our next TrustML seminar this Thursday, Nov 11 at 12pm ET 🥳🎉 Register here: us02web.zoom.us/webinar/regist… Check this thread for the speaker and talk details! 👇

Join us Nov 16 at 4pm PT for a Pinterest Labs Tech Talk! Our guest speaker, Stefanie Jegelka, will be giving a talk on Learning and Generalization in Graph Neural Networks. RSVP here bit.ly/3mZGaXl

Authorship Attribution is critical for online safety. What are the challenges and opportunities for Authorship Attribution in the era of LLMs? Can we leverage LLMs to identify authorship? Check out our two recent works accepted to #SIGKDDExplorations and #EMNLP2024 findings.

🔥New #ICLR2025 paper: "𝐂𝐚𝐧 𝐊𝐧𝐨𝐰𝐥𝐞𝐝𝐠𝐞 𝐄𝐝𝐢𝐭𝐢𝐧𝐠 𝐑𝐞𝐚𝐥𝐥𝐲 𝐂𝐨𝐫𝐫𝐞𝐜𝐭 𝐇𝐚𝐥𝐥𝐮𝐜𝐢𝐧𝐚𝐭𝐢𝐨𝐧𝐬?" Knowledge editing aims to fix false facts in LLMs—but does it actually work? 🤔 We provide new insights!✨ 🔗paper & code: llm-editing.github.io

Huge congrats to Junfeng! JUNFENG FANG Excited to see the direction of Knowledge Editing being recognized! Also check out our #ICLR2025 paper "𝐂𝐚𝐧 𝐊𝐧𝐨𝐰𝐥𝐞𝐝𝐠𝐞 𝐄𝐝𝐢𝐭𝐢𝐧𝐠 𝐑𝐞𝐚𝐥𝐥𝐲 𝐂𝐨𝐫𝐫𝐞𝐜𝐭 𝐇𝐚𝐥𝐥𝐮𝐜𝐢𝐧𝐚𝐭𝐢𝐨𝐧𝐬?" (llm-editing.github.io) if you

🚀 Although #ICLR2025 is wrapping up, we’re just getting started! Join us for the 🔥𝐑𝐞𝐚𝐬𝐨𝐧𝐢𝐧𝐠 𝐚𝐧𝐝 𝐏𝐥𝐚𝐧𝐧𝐢𝐧𝐠 𝐟𝐨𝐫 𝐋𝐋𝐌𝐬 𝐖𝐨𝐫𝐤𝐬𝐡𝐨𝐩🔥 — all day Monday, April 28! 🎤 Keynotes & panels from distinguished researchers at Meta, OpenAI, DeepMind, Oxford,

🔥Today is submission deadline (June 6th AoE) of our 𝐀𝐂𝐋 𝟐𝟎𝟐𝟓 𝐰𝐨𝐫𝐤𝐬𝐡𝐨𝐩: 𝐓𝐨𝐰𝐚𝐫𝐝𝐬 𝐊𝐧𝐨𝐰𝐥𝐞𝐝𝐠𝐞𝐚𝐛𝐥𝐞 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐌𝐨𝐝𝐞𝐥𝐬!🔥 🔗workshop website: knowledgeable-lm.github.io 🔗 submission portal: openreview.net/group?id=aclwe… 🏆 The Best Paper

Excited to visit UC Berkeley this summer, hosted by Prof. Dawn Song. Happy to catch up if you are around!