Viewing profile — tMcGrath

Yes - we'd never normally turn features up this much as it breaks the model quite badly, but we put this in the post to show what that looked like in practice.

Thank you! I think some of the features we have like conditional steering make SAEs a lot more convenient to use. It also makes using models a lot more like conventional programmin…

We'll be open-sourcing these SAEs so you're not required to do this if you'd rather self-host.

We're planning to release the weights once we do a moderation pass. Our SAE was trained on LMSys (you can see this in our accompanying post: https://www.goodfire.ai/papers/mapping-…

I'm one of the authors of this paper - happy to answer any questions you might have.

The paper is measuring I/O behaviour, rather than the complexity of the mechanisms generating that behaviour. Transistors might have quite complex physics, but are designed to have…

I'm not sure that's all there is to it - at least some scientists are in it for the explanation: they want to know more about something. In this case the prediction is just a usefu…

Modern Classical Physics is a great book - I've recently started working through it (just about to move on to chapter 2). I'd be interested in chatting about it with others/cross-c…

As far as I know, a good differential-geometric understanding of nonequilibrium thermodynamics still hasn't been achieved. The central issue is understanding how changes in control…

As I understand it her point it that what we take to be fundamental particles have certain properties - charge, spin, etc - and no others. If they had other properties, this would …

I think the standard reference is probably Spivak's 'Calculus on Manifolds' but this never really did it for me. If you have a background in physics then some combination of Nakaha…

In this case that's not true though - the article is about someone who used critical thinking/reasoning skills to make a huge amount of money, and credits this in large part to his…

I'd like to flag up this paper on HMC and minibatch inference: https://arxiv.org/abs/1502.01510v1 . It seems that SGHMC might not always behave as well as you might like. I should …

I agree - it's a surprising and cool paper. There has been some work done on fooling network ensembles by constructing constructing a Bayesian posterior over weights using dropout …

You are absolutely right that a lot of work to date (especially in applications and metrics-driven research) has focused on 'functions for vision problems'. But this is far from th…

As commenters elsewhere have pointed out, this really only caps salary at around £60k - this is actually pretty low for an experienced ML engineer. I've heard Prof Pantic advocate …

I suppose it's limited by the size of 'the set of things you can do with functions'. This seems like a large enough set that I wouldn't be worried about hitting its limits too soon…

There's a reason the axes are left unlabeled in this graph. If you look at a paper that gives values over time for these quantities you'll find that this 'effect' mounts up to abou…

There's also an interesting set of applications in statistical physics: many systems can be modelled by Markov chains whose rate parameters can change through time. Changing these …

For statistics I can really recommend 'All of Statistics' by Larry Wasserman ( http://www.stat.cmu.edu/~larry/all-of-statistics/ ) It runs through a lot of important topics, partic…

If you want to approach physics from a mathematical background I strongly recommend Peter Szekeres' book 'A Course in Modern Mathematical Physics' http://www.amazon.co.uk/gp/aw/d/0…

I've been reading this and it's fantastic - approaching the subject from the tools needed to understand the way the material works mathematically (Lagrangians etc) rather than as a…