tMcGrath

5 ms·

1.
by tMcGrath 2y ago
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.
2.
by tMcGrath 2y ago
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 programming. For example, when the model is 'thinking' x, or
3.
by tMcGrath 2y ago
We'll be open-sourcing these SAEs so you're not required to do this if you'd rather self-host.
4.
by tMcGrath 2y ago
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-latent-spaces-llama/ ). No im
5.
by tMcGrath 2y ago
I'm one of the authors of this paper - happy to answer any questions you might have.
6.
Look Inside Your AI(preview.goodfire.ai)
1 points by tMcGrath 2y ago|0 comments
7.
An empirical approach to AI safety(banburismus.substack.com)
2 points by tMcGrath 3y ago|0 comments
8.
Acquisition of Chess Knowledge in AlphaZero(arxiv.org)
2 points by tMcGrath 5y ago|0 comments
9.
by tMcGrath 5y ago
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 relatively simple I/O behaviour.
10.
by tMcGrath 6y ago
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 useful check that their proposed explanation is compati
11.
by tMcGrath 8y ago
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-checking solutions. Anyone who's interest
12.
by tMcGrath 8y ago
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 parameters (for instance concentrations of catalysts i
13.
by tMcGrath 8y ago
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 lead to measurably different outcomes in collision experimen
14.
by tMcGrath 8y ago
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 Nakahara's 'Geometry, Topology and Physic
15.
by tMcGrath 9y ago
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 training in philosophy. It's possible, of co
16.
by tMcGrath 9y ago
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 say this isn't to pour cold w
17.
by tMcGrath 9y ago
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 [0]. This is an ensemble of weights for the same networ
18.
by tMcGrath 9y ago
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 the only idea - that's the point I'm tryin
19.
by tMcGrath 9y ago
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 for salary caps for PhD students leaving academia befor
20.
by tMcGrath 9y ago
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...
21.
by tMcGrath 9y ago
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 about 3 IQ points over the range discussed i
22.
by tMcGrath 10y ago
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 parameters non-quasistatically (i.e. so that the system
23.
by tMcGrath 13y ago
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, particularly in inference, with
24.
by tMcGrath 13y ago
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/0521829607/ref=redir_mdp_mob...
25.
by tMcGrath 13y ago
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 series of separate physical problems is an excellent idea.
26.
by tMcGrath 14y ago
I'm afraid I have to disagree here. As someone who runs a small business which, for not-immediately-relevant reasons isn't in a position to try for startup-style explosive growth, I'd be really interested in hearing about how people in simi