Nick Bishop

Postdoctoral researcher at the University of Oxford applying ML to ABMs!

[Nick Bishop @NickBishop5]

It is a shame that what was such a good platform for academic discourse has been slowly eroded with selfish and immoral goals in mind. I have tried to follow everyone who has accounts on bky but let me know if I missed you!

[Nick Bishop @NickBishop5]

The recent comments by Elon towards my own country (the UK), let alone the rampant and harmful use of Grok on this platform is unacceptable. As a result I am moving to bluesky (https://t.co/zaLE5SogRW) 1/2

[Bob West @cervisiarius]

🚨 Postdoc available in my lab @EPFL on safe AI, alignment, LLMs, NLP, mech interp

🎯 Safe AI that truly cares about humans! No “lipstick-on-a-pig alignment” 🐖💄 after pretraining—let’s “raise” models aligned from token 1 onward 🍼

👉 Info & app:
https://t.co/qGFmZOAtHp

[Nick Bishop @NickBishop5]

Important read...

https://t.co/wRfA4vEC7H

[Lancelot Da Costa @lancelotdacosta]

Looking for prospective students & collaborators to work towards *An Agent Foundation Model*! 👨‍💻👩‍💻
Keywords: Active model discovery, active learning & model based planning.
More details:
https://t.co/vEOQOzVAls (bottom of page).
If you’re interested, get in touch! 💡

[Jamie Carragher @Carra23]

Lads, it’s Man United!

[Nick Bishop @NickBishop5]

@roydanroy I think a natural step is to limit the number of first author submissions one is allowed to submit. The focus should be on producing a single piece of high quality work - not tweets about "10 papers accepted".

[Long Tran-Thanh @ltt_hal]

Wow, Turing Award 2024 for Reinforcement Learning. A long awaited and deserved recognition:

https://t.co/JlRMNlu1oo

[FMZennaro @FMZennaro]

New work on causal abstraction by Willem Schooltink: looking at how graphical abstraction and functional abstraction are related! https://t.co/ygNuEvfLS0

[Nick Bishop @NickBishop5]

@avt_im My question was more about settings like in
https://t.co/WFnkUk1skL

where there is an explicit cost for pulling an arm rather than a theoretical one constructed for the purpose of analysis. However I guess this setting may be fundamentally different.

[Nick Bishop @NickBishop5]

@avt_im I guess you have already thought about this given that you mention UCB, but it seems like Pandora's box is very similar to a budgeted bandit problem with a simple regret benchmark. Is there a rigorous connection to be made there? Either way definitely giving the paper a read!

[Joel Dyer @JoelNMDyer]

Our BlackBIRDS python package implements inference methods for differentiable simulators in PyTorch, and features heavily in the tutorial.

You can play with it yourself by installing from https://t.co/AUHARXQdqB, and check out an example use case at https://t.co/W7x7V3riNg

[Joel Dyer @JoelNMDyer]

This Friday at #ICAIF 2024, together with @NickBishop5 and #AniCalinescu, we'll give a tutorial on building and calibrating differentiable agent-based models of financial systems!

Notebooks and materials are freely available on our tutorial website: https://t.co/bmp8UeOUJL

[Gergely Neu @neu_rips]

PLS SHARE:
I'm hiring a PhD student to work on ML theory, to begin in Fall 2025.
Topics include: generalization bounds & statistical inference via online prediction, representation learning via optimal transport, sequential decision making...
More info:
https://t.co/QFLEqWZORj

[Nick Bishop @NickBishop5]

@aryehazan I am now working in a different area and it's been a while since I have considered this problem but I would be happy to discuss further!

[Nick Bishop @NickBishop5]

@aryehazan My methods have since been hugely improved upon by those better educated in cvx opt. But methods still rely heavily on the Euclid. norm. I was looking to generalise my results to the L1 norm...but this escaped me during my PhD on both the optimisation and generalisation side.

[Tom Ashbyトム アシュビー @tomaashby]

The Simpsons on PhD students - almost twenty years old & more cutting than ever

[Nick Bishop @NickBishop5]

@rmkubinec Being Bayesian is not in the support of my prior

[Gergely Neu @neu_rips]

i always thought that physicists were weird, but i had no idea they had so little respect for their own field. wow