What could the future of global AI governance look like? 🌐 In our latest white paper, we explore theoretical options for international institutions to help harness the benefits and manage the risks of advanced artificial intelligence. Find out more: dpmd.ai/46LXUdR

🍍 I’ve just presented my work at #ICML #IMLH workshop! If you would like to discuss more details and chat about AI explanations, please come to the poster session! (Ballroom C) 🍍

Excited to announce Med-Flamingo, a new multimodal few-shot learner specialized for the medical domain! Last week, we uploaded the pre-print, now it’s finally live! Paper: arxiv.org/abs/2307.15189 Code: github.com/snap-stanford/… Model: huggingface.co/med-flamingo/m… A short 🧵. 1/n

Excited to share our perspective on AI + Science! We review how AI accelerates the scientific inquiry loop (hypothesis -> experiments -> observations) AI for Science

Do Machine Learning Models Memorize or Generalize? pair.withgoogle.com/explorables/gr… An interactive introduction to grokking and mechanistic interpretability w/ Asma Ghandeharioun, nada hussein, Nithum, Martin Wattenberg and iislucas (Lucas Dixon)

Re: #ICML2023 🚀 I want to give a big shoutout to both old and new friends who made #ICML2023 experience incredible! 🌴 I am grateful for the Travel Award I received from the ICML #IMLH workshop, where I presented my paper! Thanks to #WiML for their support as well!

How might we craft an artificial society that reflects human behavior? My paper, which introduced “generative agents,” will be presented at #UIST2023 and now has an open-source repo! w/ Joey O'Brien Carrie Cai Meredith Ringel Morris Percy Liang Michael Bernstein arxiv.org/abs/2304.03442 🧵

📢 Exciting News! Our latest paper is now out in Nature Biotech 🌱🧬 We developed GEARS---an AI method to predict cellular responses to genetic perturbation. 🧪🔬 🔗 Link to the paper: nature.com/articles/s4158… 🧬 Unraveling genetic interactions in cancer, regenerative medicine,

Excited to share the Stanford Estimator of Electrocardiogram Risk (SEER), an AI risk score based on the ECG, out today in npj Digital Medicine with Marco Perez James Zou euan ashley David Ouyang, MD Pierre Elias, MD and many others! StanfordDBDS Cedars-Sinai Academic Medicine Columbia Cardiology Fellows 1/n

How do existing knowledge, technologies, and power structures reinforce each other in Natural Language Processing research? We find that involvement in NLP research becomes more unequal as we move from unlabeled data to deployed systems. williamheld.com/coloniality

📢 I'm on the faculty job market for 2024! 📢 Grateful for any RTs and pointers! My vision is to develop biomedical AI systems that exhibit generalist capabilities, see e.g. nature.com/articles/s4158… Below a selected overview of my prior & ongoing work:

We're looking for a postdoc fellow to work on human-AI interaction around civil discourse 👉 visit our job posting and apply by Feb 15: digitaleconomy.stanford.edu/about/job-posi… #NLProc

Stop by our posters today to hear about #visualsemiotics #dialects #caricature 🔥 #EMNLP2023

Can deep learning work on small data with far more features than samples? We present PLATO: a method that achieves the state-of-the-art on such datasets by using prior domain information! neurips.cc/virtual/2022/p… 🧵 Published in #NeurIPS2023 with Hongyu Ren @kexinhuang5 Jure Leskovec

Presenting this spotlight poster at NeurIPS today! Come talk to us about systematic, automated evaluations and theory of mind!! #539 Great Hall & Hall B1+B2 (level 1) Thu 14 Dec 5 p.m. CST — 7 p.m. CST neurips.cc/virtual/2023/p… See you there!

Aligned LLMs should be helpful, harmless, and adopt user preferences. But whose preferences are we aligning to and what are unintended effects on global representation? We find SFT and Preference Tuning steer LLMs towards US English use and opinions. arxiv.org/abs/2402.15018 🧵

Before taking it out on your roommate for leaving dirty dishes out, you probably want to practice your conflict resolution skills first. Expert conflict resolution trainers, however, are EXPENSIVE. What if we practiced with an angry LLM instead? 😈 #CHI24 arxiv.org/abs/2309.12309

So excited to share our new preprint, “LLMs generate structurally realistic social networks but overestimate political homophily”! Joint work co-led by Alicja Chaszczewicz, along with Emma Wang, Maya Josifovska, Emma Pierson, and Jure Leskovec Paper: arxiv.org/abs/2408.16629