hylandsl.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

found a copy of the NeurIPS 1989 proceedings at the office

submitted 20 days ago • 0 comments

One of my grand interpretability goals is to improve human scientific understanding by analyzing scientific discovery models, but this is the most convincing case yet that we CAN learn from model interpretation: Chess grandmasters learned new play concepts from AlphaZero's internal representations.

submitted 30 days ago • 2 comments

I genuinely think about this book all the time

submitted 42 days ago • 0 comments

Happy to announce a collaboration with the Mayo Clinic to advance our research in radiology report generation! newsnetwork.mayoclinic.org/discussion/m... Tagging some of the core team: @valesalvatelli.bsky.social @fepegar.com @maxilse.bsky.social @sambondtaylor.bsky.social @anton-sc.bsky.social

submitted 42 days ago • 1 comment

me, eating potato waffles straight from the toaster: see, vegan diets are healthier

submitted 51 days ago • 0 comments

enjoying listening to the BBC Reith Lectures (www.bbc.co.uk/programmes/b...). In the 2021 edition, Stuart Russell offhandedly predicted AlphaFold winning the Nobel prize!

submitted 52 days ago • 0 comments

Working on a "special" kind of data (health; medical images etc.) means experiencing the disappointment of realising the method in a paper relies critically on "having a very good image captioner", "asking GPT-4V", etc.

submitted 54 days ago • 2 comments

I love having a very specific question I'm trying to answer and just going to town on semantic scholar.

submitted 54 days ago • 1 comment

Let us start 2025 in a positive mood: here are 10 methods things researchers can worry less about in 2025

submitted 65 days ago • 15 comments

THREAD This Is The Story Of The The Pernicious Rise of AI-Generated Papers and their Online Impact An Incomplete History Told In The Voice of Documentarian Adam Curtis 1/31

submitted 69 days ago • 29 comments

feeling a but under the weather this week … thus an increased level of activity on social media and blog: kyunghyuncho.me/i-sensed-anx...

submitted 67 days ago • 19 comments

A word of caution: it might feel like that grant/paper/review cannot possibly wait over the Christmas period, but you can't get time with friends and family back. You likely won't be like "really wish I'd submitted that" in 5 years time but you might regret not spending time with those you love.

submitted 69 days ago • 3 comments

Microsoft's Computational Social Science group may have the opportunity to hire one researcher Senior: 0-3 yrs post PhD jobs.careers.microsoft.com/global/en/jo... Principal: 3+ yrs post PhD jobs.careers.microsoft.com/global/en/sh... Please note: our ability to hire this season is not certain

submitted 69 days ago • 6 comments

Isaac Kohane highlighting that AI is already being used by health insurance companies to deny claims #ML4H

submitted 72 days ago • 2 comments

Simulating a skin lesion dataset for evaluating dermatology models, from Adarsh Subbaswamy at ML4H

submitted 72 days ago • 0 comments

“Most people who have died from malaria were never in an EHR system” - Megan Coffee on the panel (“The Promise of Foundation Models in Healthcare”) at ML4H

submitted 73 days ago • 0 comments

Max Tegmark pointing out the benefits of “tool AI” over “AGI” at the #NeurIPS2024 Safe Generative AI workshop

submitted 73 days ago • 4 comments

my roller derby name is Kernel Panic 💅

submitted 73 days ago • 0 comments

It’s kicking off in the #NeurIPS2024 town hall slido

submitted 75 days ago • 1 comment

Rosalind Picard on involving the real people who will be impacted/benefit from your research in your research. (One of many excellent lessons from her talk on “How to optimise what matters most?”) #NeurIPS2024

submitted 75 days ago • 1 comment

Lidong Zhou giving lessons on scaling from the systems community #NeurIPS2024

submitted 76 days ago • 0 comments

“How would you describe your research?” #NeurIPS2024

submitted 77 days ago • 1 comment

not enough people are talking about the concerning proliferation of radar charts in machine learning research towardsdatascience.com/ghosts-on-th...

submitted 77 days ago • 1 comment

whenever someone presents a “here’s the history of this field” slide I can’t help but notice an absence of women

submitted 77 days ago • 1 comment

profound questions in the #NeurIPS2024 slido

submitted 78 days ago • 0 comments

Nice to see machine learning for health in the top 20 topics in NeurIPS papers #NeurIPS2024

submitted 78 days ago • 1 comment

My group is hiring postdoc(s) and I will be at #NeurIPS2024 if you want to talk - send me a DM or find me on Whova! UK-based, 2 year position. We work on AI/ML for health, specifically building and understanding/explaining deep (multimodal) models on healthcare data, especially medical imaging.

submitted 79 days ago • 0 comments

Logged into Whova for the first time since 2019. It's almost time for #NeurIPS2024!

submitted 81 days ago • 0 comments

When I will respond to your email

submitted 98 days ago • 41 comments