Ellis Unit
Ellis Unit
  • 50
  • 25 957
Cambridge Seminar Series Alex Tong
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for September 2024 will be Alexander Tong. Details of his talk can be found below.
Abstract:
In this talk I will first discuss how optimal transport can be combined with flow matching for improved efficiency in training and inference. By reducing variance in the objective and curvature in the trajectories. I will then discuss how these methods can be applied towards protein design tasks with the FoldFlow family of models.
Переглядів: 173

Відео

Cambridge Ellis Seminar series - Simon Olsson - 15 November 2024
Переглядів 152Місяць тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for November 2024 will be Simon Olsson. Details of his talk can be found below. Title: “Generative models as efficient surrogates for molecular dynamics simulations” Abstract: Molecular dynamics (MD) is an important simulation technique in the natural sciences and ...
Cambridge ELLIS Seminar Speaker Gabriel Synnaeve
Переглядів 1475 місяців тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for July 2024 will be Gabriel Synnaeve. Details of his talk can be found below. Title: “Multi-token Prediction and Exploring LM Losses“ Abstract: Large language models such as GPT and Llama are trained with a next-token prediction loss. In this work, we suggest tha...
ELLIS Seminar Series Prof Po Ling Loh Differentially private M estimation via noisy optimization
Переглядів 868 місяців тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for Apr 2024 will be Prof. Po-Ling Loh. Details of her talk can be found below. Title: “Differentially private M-estimation via noisy optimization" Abstract: We present a noisy composite gradient descent algorithm for differentially private statistical estimation i...
ELLIS Seminar Series - Dr. Rika Antonova
Переглядів 13710 місяців тому
Title: Enabling Self-sufficient Robot Learning Abstract: Autonomous exploration and data-efficient learning are important ingredients for helping machine learning handle the complexity and variety of real-world interactions. In this talk, I will describe methods that provide these ingredients and serve as building blocks for enabling self-sufficient robot learning. First, I will outline a famil...
Cambridge ELLIS Seminar Series - Dr Tian Xie
Переглядів 28011 місяців тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for 2024 will be Dr. Tian Xie. Details of his talk can be found below. Title: "a generative model for inorganic materials design" Abstract: The design of functional materials with desired properties is essential in driving technological advances in areas like energ...
Halting Climate Change- Carl Edward Rasmussen- NeurIPS@Cambridge- 8 Dec 2023
Переглядів 614Рік тому
Halting Climate Change- Carl Edward Rasmussen- NeurIPS@Cambridge- 8 Dec 2023
Cambridge ELLIS Seminar Series Prof Gustau Camps Valls 28 Nov 23 13:30 SD 480p
Переглядів 115Рік тому
Cambridge ELLIS Seminar Series- Prof. Gustau Camps-Valls- 28 Nov 23- 13:30
Cambridge ELLIS Seminar Series Dr David Barber SD 480p
Переглядів 72Рік тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Our next speaker for 2023 will be Dr. David Barber. Details of his talk can be found below. Title: “Training Energy Based Models “ Abstract: Energy Based models are undirected models of the formp(x) \propto exp(-f(x)). These are very general models and provide an alternative to the...
Cambridge ELLIS Seminar Series Dr Jasper Snoek
Переглядів 129Рік тому
Title: “Kernel (GP) Regression with Infinite-Width Neural Networks on Millions of Examples “ Abstract: I’ll talk about some of our recent work on scaling up Gaussian process regression with expressive kernels on structured data. Neural kernels have drastically increased performance on diverse and nonstandard data modalities but require significantly more compute, which previously limited their ...
Cambridge ELLIS Seminar Series Petar Veličković SD 480p
Переглядів 169Рік тому
The Cambridge ELLIS Unit Seminar Series holds talks by leading researchers in the area of machine learning and AI. Next first speaker of 2023 will be Dr. Petar Veličković. Details of his talk can be found below. Title: “Reasoning Algorithmically: from Toy Experiments to AGI Modules” Abstract: Neural networks that are able to reliably execute algorithmic computation may hold transformative poten...
Cambridge Ellis Unit Seminar Series - Dr Manuel Gomez Rodriguez - 31 Jan 2023- 2pm
Переглядів 171Рік тому
Cambridge Ellis Unit Seminar Series - Dr Manuel Gomez Rodriguez - 31 Jan 2023- 2pm
Cambridge ELLIS Seminar Series Zeynep Akata
Переглядів 3412 роки тому
Cambridge ELLIS Seminar Series Zeynep Akata
Andrew Thomas To reify or not to reify the graph
Переглядів 482 роки тому
Andrew Thomas To reify or not to reify the graph
Bob Carpenter Comprehension, maps, and partial evaluation in differentiable programming with appli
Переглядів 902 роки тому
Bob Carpenter Comprehension, maps, and partial evaluation in differentiable programming with appli
Chris Rackauckas Integrating solvers w/ probabilistic programming through differentiable programming
Переглядів 1,3 тис.2 роки тому
Chris Rackauckas Integrating solvers w/ probabilistic programming through differentiable programming
Sharad Vikram Jaxprs dead simple, surprisingly versatile
Переглядів 2232 роки тому
Sharad Vikram Jaxprs dead simple, surprisingly versatile
Zichun Ye MindIR the intermediate representation of MindSpore
Переглядів 632 роки тому
Zichun Ye MindIR the intermediate representation of MindSpore
Maria Gorinova Program Analysis of Probabilistic Programs
Переглядів 1922 роки тому
Maria Gorinova Program Analysis of Probabilistic Programs
Cambridge ELLIS Seminar Series Maria Perez Ortiz SD 480p
Переглядів 1422 роки тому
Cambridge ELLIS Seminar Series Maria Perez Ortiz SD 480p
Carl Henrik Ek Bayesian Non Parametrics P2
Переглядів 4292 роки тому
Carl Henrik Ek Bayesian Non Parametrics P2
Carl Henrik Ek Bayesian Non Parametrics P1
Переглядів 2532 роки тому
Carl Henrik Ek Bayesian Non Parametrics P1
Cecily Morrison Human Centric AI: The Opportunity of Teachable AI Systems
Переглядів 602 роки тому
Cecily Morrison Human Centric AI: The Opportunity of Teachable AI Systems
Tom Rainforth Bayesian Experimental Design and Active Learning P2
Переглядів 1,2 тис.2 роки тому
Tom Rainforth Bayesian Experimental Design and Active Learning P2
Tom Rainforth Bayesian Experimental Design and Active Learning P1
Переглядів 3,2 тис.2 роки тому
Tom Rainforth Bayesian Experimental Design and Active Learning P1
Jose Miguel Hernandez Lobato The Linearized Laplace Approximation in Bayesian Neural networks P3
Переглядів 4062 роки тому
Jose Miguel Hernandez Lobato The Linearized Laplace Approximation in Bayesian Neural networks P3
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 1
Переглядів 3,9 тис.2 роки тому
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 1
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 2
Переглядів 1,7 тис.2 роки тому
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 2
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 3
Переглядів 1,1 тис.2 роки тому
Petar Veličković Graph Neural Networks: Geometric, Structural and Algorithmic Perspectives Part 3
David Krueger AI Safety and Alignment- Part 1
Переглядів 9652 роки тому
David Krueger AI Safety and Alignment- Part 1

КОМЕНТАРІ

  • @sheevys
    @sheevys Місяць тому

    How come invenia is still a sponsor. Do they still exist, after all what happened?

  • @raaviaggarwal8223
    @raaviaggarwal8223 7 місяців тому

    One issue with the proposal is that it creates incentives for country governments to keep investing fossil fuel technologies, to get more money from the club. However, the market will react and reduce carbon emissions. While this is a good thing, it is very likely that citizens will oppose such a high carbon price, in underdeveloped countries, if the government does not make transfers to people from the money it gets from the club. But, if the government gives money back to people, especially companies, people will reinvest that money in fossil fuel production and consumption. That's because the club creates an incentive to keep emitting (and hence get more money), especially in populated countries.

    • @positivedefinite
      @positivedefinite 6 місяців тому

      How does the club create an incentive to keep emitting? Every ton of emissions is a net cost to the country emitting it. It pays p but only receives back a fraction of that.

  • @mueletech
    @mueletech 9 місяців тому

    more like machine rovers

  • @bjmcculloch
    @bjmcculloch Рік тому

    Everything Chris described regarding the documentation problem matches my experience with Julia. It’s clearly a fantastic problem to have, but it needs to be addressed to promote adoption.

  • @L4ky13
    @L4ky13 Рік тому

    Horrible sound quality

  • @Alexfnplays
    @Alexfnplays 2 роки тому

    really nice and clear talk!

  • @nowithinkyouknowyourewrong8675
    @nowithinkyouknowyourewrong8675 2 роки тому

    Nice to see NP's used on real data, thanks for sharing

  • @stephen_pfrimmer
    @stephen_pfrimmer 2 роки тому

    Reinforcement Learning. Very interesting!

  • @NobodyMinus
    @NobodyMinus 2 роки тому

    Excellent, please keep posting.

  • @andreasholzinger7056
    @andreasholzinger7056 2 роки тому

    Graphs and Graph neural networks are really great - I like the talks of Petar very much and can fully reommennd it - very valuable!