Ivor Simpson
@ivorsimpson.bsky.social
Associate Professor in AI and academic lead for SussexAI. Bayesian ML for medical imaging, computer vision & environmental monitoring.
Please share with your contacts! Deadline is 19th February.
Email me and/or DSAI_administration@sussex.ac.uk if you have any questions about the process. @sussexai.bsky.social @drtnowotny.bsky.social @wijdr.bsky.social anyone else on here yet?!
Email me and/or DSAI_administration@sussex.ac.uk if you have any questions about the process. @sussexai.bsky.social @drtnowotny.bsky.social @wijdr.bsky.social anyone else on here yet?!
December 9, 2024 at 7:00 PM
Please share with your contacts! Deadline is 19th February.
Email me and/or DSAI_administration@sussex.ac.uk if you have any questions about the process. @sussexai.bsky.social @drtnowotny.bsky.social @wijdr.bsky.social anyone else on here yet?!
Email me and/or DSAI_administration@sussex.ac.uk if you have any questions about the process. @sussexai.bsky.social @drtnowotny.bsky.social @wijdr.bsky.social anyone else on here yet?!
We even had slushies afterwards...
December 9, 2024 at 5:49 PM
We even had slushies afterwards...
Amazing news, congratulations Maria!
December 4, 2024 at 5:27 PM
Amazing news, congratulations Maria!
For those that want to dig into this area, it’s worth reading a bit more about tools for understanding learning in ML. I’d definitely recommend starting with @simonprinceai.bsky.social blogs on Gradient Flow and the Neural Tangent Kernel (NTK) (links from udlbook.github.io/udlbook/)
Understanding Deep Learning
udlbook.github.io
November 29, 2024 at 3:36 PM
For those that want to dig into this area, it’s worth reading a bit more about tools for understanding learning in ML. I’d definitely recommend starting with @simonprinceai.bsky.social blogs on Gradient Flow and the Neural Tangent Kernel (NTK) (links from udlbook.github.io/udlbook/)
One of the authors wrote a very nice thread on this, so I’ll point to that rather than try and explain the methodology myself! bsky.app/profile/alic...
From double descent to grokking, deep learning sometimes works in unpredictable ways.. or does it?
For NeurIPS(my final PhD paper!), @alanjeffares.bsky.social & I explored if&how smart linearisation can help us better understand&predict numerous odd deep learning phenomena — and learned a lot..🧵1/n
For NeurIPS(my final PhD paper!), @alanjeffares.bsky.social & I explored if&how smart linearisation can help us better understand&predict numerous odd deep learning phenomena — and learned a lot..🧵1/n
November 29, 2024 at 3:36 PM
One of the authors wrote a very nice thread on this, so I’ll point to that rather than try and explain the methodology myself! bsky.app/profile/alic...
They found a divergence in the effective complexity of the model when run on the testing, rather than the training set, seemed to be linked to model generalisation. My interpretation of this is the model can use memorisation on training examples, but needs to interpolate on test examples.
November 29, 2024 at 3:36 PM
They found a divergence in the effective complexity of the model when run on the testing, rather than the training set, seemed to be linked to model generalisation. My interpretation of this is the model can use memorisation on training examples, but needs to interpolate on test examples.
In our new "Advanced Methods for Machine Learning" module, this week seminar dug into an upcoming NeurIPS paper arxiv.org/abs/2411.00247 that provides a new tool for analysing changes in effective model complexity when predicting on the training/test set over the course of training.
Deep Learning Through A Telescoping Lens: A Simple Model Provides Empirical Insights On Grokking, Gradient Boosting & Beyond
Deep learning sometimes appears to work in unexpected ways. In pursuit of a deeper understanding of its surprising behaviors, we investigate the utility of a simple yet accurate model of a trained neu...
arxiv.org
November 29, 2024 at 3:36 PM
In our new "Advanced Methods for Machine Learning" module, this week seminar dug into an upcoming NeurIPS paper arxiv.org/abs/2411.00247 that provides a new tool for analysing changes in effective model complexity when predicting on the training/test set over the course of training.