x

LESSWRONG
LW

Stephen Fowler — LessWrong

Stephen Fowler

Top postsTop post

Stephen Fowler

Message

Masters student in Physics at the University of Queensland.

I am interested in Quantum Computing, physical AI Safety guarantees and alignment techniques that will work beyond contemporary models.

2796

Ω

24

13

225

4y

Stephen Fowler

Masters student in Physics at the University of Queensland.

I am interested in Quantum Computing, physical AI Safety guarantees and alignment techniques that will work beyond contemporary models.

Top postsTop post

Race Along Rashomon Ridge

Produced As Part Of The SERI ML Alignment Theory Scholars Program 2022 Research Sprint Under John Wentworth Two Deep Neural Networks with wildly different parameters can produce equally good results. Not only can a tweak to parameters leave performance unchanged, but in many cases, two neural networks with completely different weights and biases produce identical outputs for any input. The motivating question: Given two optimal models in a neural network's weight space, is it possible to find a path between them comprised entirely of other optimal models? In other words, can we find a continuous path of tweaks from the first optimal model to the second without reducing performance at any point in the process? Ultimately, we hope that the study of equivalently optimal models would lead to advances in interpretability: for example, by producing models that are simultaneously optimal and interpretable. Introduction Your friend has invited you to go hiking in the Swiss Alps. Despite extensive planning, you completely failed to clarify which mountain you were supposed to climb. So now you're both currently standing on different mountaintops Communicating via yodels, your friend requests that you hike over to the mountain he's already on; after all, he was the one who organised the trip. How should you reach your friend? A straight line path would certainly work, but it's such a hassle. Hiking all the way back down and then up another mountain? Madness. Looking at your map, you notice there's a long ridge that leaves your mountain at roughly the same height. Does this ridge lead to reach your friend's mountain and is it possible to follow this ridge to reach your friend? Similarly, we can envision two optimal models in a neural net's weight space that share a "ridge" in the loss landscape. There is a natural intuition that we should be able to "walk" between these two points without leaving the ridge. Does this intuition hold up? And what of two neural net

Searching for Modularity in Large Language Models

Scaffolded LLMs: Less Obvious Concerns

Broad Basins and Data Compression

To what extent is the UK Government's recent AI Safety push entirely due to Rishi Sunak?

And hence, to what extent might we expect all that momentum to suddenly dissipate? In other words, if Sunak was to suddenly leave office (as is tradition for British PMs) would you expect the Frontier AI Taskforce to continue receiving funds, or would it be quietly shelved. I get the...

Oct 27, 2023•23

What are the best published papers from outside the alignment community that are relevant to Agent Foundations?

I am after papers that you have stumbled across that may be relevant to the core of the understanding the existential AI alignment problem despite not having explicit links to AI alignment. Here are papers I have found that fit the above criteria to give you a better idea of...

Aug 5, 2023•20

Scaffolded LLMs: Less Obvious Concerns

“Words, like tranquil waters behind a dam, can become reckless and uncontrollable torrents of destruction when released without caution and wisdom.” — William Arthur Ward In this post I aim to shed light on lesser-discussed concerns surrounding Scaffolded LLMs (S-LLMs). The core of this post consists of three self-contained discussions,...

Jun 16, 2023•34

What do beneficial TDT trades for humanity concretely look like?

Who do you think you're trading with and why? What explicitly is being traded? What mechanism enables both parties to trust the other? I am quite pessimistic TDT related trades. Not because I think it isn't a real thing, but because I think the kinds of "trade offers" we encounter...

Jun 10, 2023•4

Requisite Variety

Epistemics: First two sections are almost half a century old ideas from early Control Theory. Last section on alignment is speculative. The Law of Requisite Variety Illustration inspired by Norman and Bar-Yam, 2018 TL;DR: Requisite Variety concerns the straightforward relationship between the number of actions we can take and the...

Apr 21, 2023•6

Ng and LeCun on the 6-Month Pause (Transcript)

This is a transcript a brief chat between Andrew Ng and Yann LeCun on the prospect of the 6-Month Pause. They are both highly skeptical of the pause and more generally of "AI doomerism". Yann also briefly touches on why he thinks LLMs won't scale to AGI. Full video can...

Apr 9, 2023•29

No Summer Harvest: Why AI Development Won't Pause

Cleo Nardo recently highlighted Yudkowsky and others discussing an "AI Summer Harvest" (or simply "Summer Harvest"). This potential near-future era would involve a pause in research on state-of-the-art (SOTA) models, enabling us to fully enjoy the economic benefits of the extraordinary technology we currently possess. There would be a widespread...

Apr 6, 2023•14

Load More (7/17)