user avatar
yingzhen
@liyzhen2
teaching machines 🤖 to learn 🔍 and fantasize 🪄 now 🇬🇧 @ImperialCollege @ICComputing 🇸🇬 @NTUsg @NTU_ccds ex 🇬🇧 @MSFTResearch @CambridgeMLG
yingzhenli.net
Joined April 2012
164
Following
4,432
Followers
Posts
  • user avatar
    yingzhen
    @liyzhen2
    Jul 30, 2025
    I read the first 4 books extensively during my PhD, highly recommended 👍 I'd also highlight the 5th book as my first read re deep learning. Mind-blowing for a young math undergrad (me) at the time, made me decide to go for ML
    user avatar
    Anthony Bonato
    @Anthony_Bonato
    Jul 27, 2025
    What four math books had a big influence on your mathematical thinking? I'll start:
    134K
  • user avatar
    yingzhen
    @liyzhen2
    Aug 12, 2022
    Introducing NeurIPS 2022 workshop on score-based methods: score-based-methods-workshop.github.io Generative NNs using ∇log p(x) are amazing🥰. Even more exciting that research on the magic score function is way beyond that😍, and we hope to connect ML, stats & app. researchers here!
  • user avatar
    yingzhen
    @liyzhen2
    Dec 6, 2020
    We (Cheng Zhang and myself) are thrilled to share with you some recent advances in approximate inference (😇≈😈)@NeurIPSConf #NeurIPS2020 No registration needed 🥳 nips.cc/virtual/2020/p… Will cover the basics, some advances 👇, and of course applications 😊
  • user avatar
    yingzhen
    @liyzhen2
    Nov 3, 2025
    The VCL paper has arguably the first example of modern continual learning for GenAI: VAEs trained on digit/alphabet images 1-by-1 arxiv.org/abs/1710.10628 Coded by yours truly ☺️ who was (and still is) 🥰 in generative models. Time to get back to continual learning again?
    18K
  • user avatar
    yingzhen
    @liyzhen2
    Jan 20, 2023
    Proud advisor moment: 1st PhD student’s 1st research project accepted at #ICLR2023 on 1st try of submission 😆 a dream come true: seeing big elephants in Africa😍
    33K
  • user avatar
    yingzhen
    @liyzhen2
    Jan 19, 2025
    Tons of papers re diffusion/flow matching at ML confs these days, but to my surprise very few of them consider learning the prior🤔 Am I missing any important work here? 🙏 for suggestions
    70K
  • user avatar
    yingzhen
    @liyzhen2
    Jan 8, 2021
    This week I started my new job at @ICComputing as a lecturer (US assist prof), WFH in Cambridge still but super excited😆 Thanks for everyone I've met for the amazing past, stay safe and I'm sure we will meet again in the future🥰
  • user avatar
    yingzhen
    @liyzhen2
    Aug 12, 2020
    Cheng Zhang and myself will give a tutorial on approximate inference at #neurips2020, very excited!😆 Wow, feels like it was only yesterday that I published my first ML paper at NeurIPS 2015 on approximate inference (and started computing integrals all the time)😅
    user avatar
    Danielle Belgrave
    @DaniCMBelg
    Aug 12, 2020
    Thanks to everyone who submitted tutorial proposals for this year's @NeurIPSConf. It was amazing to see all the great work that went in to putting together these proposals. We are happy to announce this year's tutorials at #neurips2020 medium.com/@NeurIPSConf/t…
  • user avatar
    yingzhen
    @liyzhen2
    Sep 5, 2025
    We show how to make LLM in-context learning approximately Bayesian & decompose uncertainty IMO this is proper approximate inference 🥰 applied to LLMs Led by awesome students @shavindra_j @jacobyhsi88 Filippo & Wenlong 👍 Example👇by prompting, bandits & NLP examples in paper
    user avatar
    Statistics (Machine Learning) Papers
    @StatsPapers
    Sep 4, 2025
    Automated by @PremiumAccts
    Variational Uncertainty Decomposition for In-Context Learning. arxiv.org/abs/2509.02327
    21K
  • user avatar
    yingzhen
    @liyzhen2
    Oct 31, 2025
    Instead of learning SDEs, we directly parameterise and learn the weak solutions of SDEs from sequence/time-series data. Training and inference done both in simulation-free manner. Kudos to @naoki_kiyohara for his great innovation 👍 see you @NeurIPSConf and @EurIPSConf 😊
    user avatar
    Naoki Kiyohara
    @naoki_kiyohara
    Oct 30, 2025
    📢 Excited to share our #NeurIPS2025 paper on Neural Stochastic Flows! We learn SDE transition distributions directly, enabling solver-free sampling across arbitrary time gaps 🌐 nkiyohara.github.io/nsf-neurips2025 📄 arxiv.org/abs/2510.25769 🙌 Thanks to: @liyzhen2 @Ed__Johns (1/5)
    00:00
    16K
  • user avatar
    yingzhen
    @liyzhen2
    Sep 22, 2022
    My lecture on intro to Bayesian neural networks at ProbAI school 2022, including hands-on tutorials
    user avatar
    Eliezer de Souza da Silva
    @sereliezer
    Sep 22, 2022
    Replying to @sereliezer
    - Bayesian Neural Networks by Yingzhen Li (@liyzhen2) youtu.be/cRzNWVjnD6I - Advanced Bayesian Neural Networks by José Miguel Hernández–Lobato (@jmhernandez233) youtube.com/watch?v=MoJVCc… - Closing of ProbAI 2022 by Luigi Acerbi (@AcerbiLuigi) youtube.com/watch?v=L13EYw…
    Colab logo
    colab.research.google.com
    ProbAI 2022 BNN Tutorial - regression.ipynb
    Colab notebook
  • user avatar
    yingzhen
    @liyzhen2
    Sep 14, 2022
    3/3 papers accepted 😆at #NeurIPS2022 on 3 different topics: - infomin learning - probabilistic models on sets - repairing neural networks will share details soon when they go public
  • user avatar
    yingzhen
    @liyzhen2
    May 5, 2025
    #AISTATS2025 day 3 keynote by Akshay Krishnamurthy about how to do theory research on inference time compute 👍 @aistats_conf
    15K
  • user avatar
    yingzhen
    @liyzhen2
    Jan 30, 2022
    Join us to discuss the latest advances in approximate inference and probabilistic models at AABI 2022 on Feb 1-2! Webinar registration: approximateinference.org/schedule/ We have an amazing line-up of speakers, panelists and papers👍 @vincefort @Tkaraletsos @s_mandt @ruqi_zhang