user avatar
Willie Neiswanger
@willieneis
Assistant Professor @USC in CS + AI. Previously @Stanford, @SCSatCMU. Machine Learning, Decision Making, AI-for-Science, Generative Models.
Los Angeles
willieneis.github.io
Joined March 2009
285
Following
1,490
Followers
Posts
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 7, 2021
    (1/9) Presenting: Bayesian Algorithm Execution (BAX) and the InfoBAX algorithm. Bayesian optimization finds global optima of expensive black-box functions. But what about other function properties? w/ @KAlexanderWang @StefanoErmon at #ICML2021 URL: willieneis.github.io/bax-website
    00:00
  • user avatar
    Willie Neiswanger
    @willieneis
    Nov 15, 2023
    Excited to share that I will join @USC as an Asst. Professor of Computer Science in Jan 2024—and I’m recruiting students for my new lab! 📣 Come work at the intersection of machine learning, decision making, generative AI, and AI-for-science. More info: willieneis.github.io/lab
    70K
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 12, 2022
    (1/4) Very excited about a new project we’ve been working on — Betty! Betty is an autodiff library for multilevel optimization and generalized meta-learning ✨ GitHub: github.com/leopard-ai/bet… ArXiv: arxiv.org/abs/2207.02849 Docs: leopard-ai.github.io/betty/ Led by @sangkeun_choe 🙌
  • user avatar
    Willie Neiswanger
    @willieneis
    Jan 8, 2023
    📢 Get ready for a deep dive into the world of modern experimental design and active learning! We’re starting a Reading Group that explores these techniques, and their applications to real-world problems. 🚨 Details: realworldml.github.io cc/ @ilijabogunovic @mutny_ml
    26K
  • user avatar
    Willie Neiswanger
    @willieneis
    Sep 27, 2023
    Reminder! Oct 4 deadline for #NeurIPS2023 Workshop on Adaptive Experimental Design & Active Learning in the Real World🔬 💸$1000 best student paper award.👥Great set of speakers: @MihaelaVDS @annadgoldie @nathankallus @eytan @EmmaBrunskill @erika_alden_d realworldml.github.io/neurips2023
    12K
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 7, 2021
    Replying to @willieneis
    (9/9) Our method has connections to stats/ML areas such as Bayesian optimal experimental design, information theory, optimal sensor placement, stepwise uncertainty reduction, likelihood-free Bayesian inference, and more. See more discussion in our paper arxiv.org/abs/2104.09460
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 18, 2022
    Check out the fantastic set of accepted papers in the ReALML at #icml2022 workshop! (Workshop on Adaptive Experimental Design and Active Learning in the Real World 🌎) realworldml.github.io/accepted/ We hope to see you there, this Friday, in Room 309!
  • user avatar
    Willie Neiswanger
    @willieneis
    Jun 2, 2022
    Final reminder: ReALML @ #icml2022 workshop (realworldml.github.io) submission deadline is at the end-of-day on June 3 (anywhere on earth) — still plenty of time to submit! Note that we accept papers submitted to NeurIPS, and there is a $1000 best student paper award!
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 7, 2021
    Replying to @willieneis
    (2/9) Methods like Bayes opt / quadrature can be viewed as estimating properties of a black-box function (e.g. global optima, integrals). But in many applications we also care about local optima, level sets, top-k optima, boundaries, integrals, roots, graph properties, and more
    Example properties of black-box functions and associated applications in which it is useful to estimate these properties.
  • user avatar
    Willie Neiswanger
    @willieneis
    Jan 31, 2023
    This week in our online REALML reading group: Emmanuel Bengio (@folinoid) will present “Introduction to GFlowNet” Thursday Feb 2 at 10am PT / 6pm GMT / 7pm CET — see you there! For more info: realworldml.github.io cc @ilijabogunovic @mutny_ml
    1.7K
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 17, 2020
    See the list of accepted papers here: realworldml.github.io/accepted/ Lots of interesting work!
    user avatar
    Yisong Yue
    @yisongyue
    Jul 17, 2020
    (Virtual) Workshop on Real World Experiment Design & Active Learning! Tune in Saturday July 18th! #ICML2020 site: icml.cc/virtual/2020/w… Talk videos will be made publicly available at a later date.
  • user avatar
    Willie Neiswanger
    @willieneis
    May 20, 2021
    Check out our upcoming ICML 2021 workshop on Machine Learning for Data (ml4data)! This workshop will focus on how ML techniques can be used to facilitate a range of data operations (e.g. ML-assisted labeling, synthesis, selection, augmentation), and the associated challenges.
    user avatar
    ml4data
    @ml4data
    May 20, 2021
    🚨Announcing the ML4data workshop at ICML 2021🚨 — a workshop focused on how we use ML for our most precious resource: data sites.google.com/view/ml4data #icml2021 #icml21 #ml4data
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 12, 2022
    Replying to @willieneis
    (4/4) Fun fact: Betty is named after @sangkeun_choe’s dog, Betty. A big thanks to our other collaborators and co-developers @ericxing and @cmuptx 🙌 To contribute: github.com/leopard-ai/bet… Happy multilevel optimization programming!
  • user avatar
    Willie Neiswanger
    @willieneis
    Jul 12, 2022
    Replying to @willieneis
    (2/4) Betty can be used for many applications: github.com/leopard-ai/bet… * Differentiable hyperparameter tuning * Data/sample reweighting * Domain adaptation in pretraining & finetuning * Differentiable architecture search (DARTS) * Implicit MAML/few shot learning * RL, GANs, ++