Sign in

David Pfau

@davidpfau.com
10K followers 548 following 953 posts

So far I have not found the science, but the numbers keep on circling me. Views my own, unfortunately.

PostsRepliesMedia
David Pfau @davidpfau.com · 22/12/2025
Rather than write down a solution through a stroke of genius, we trained a Transformer-like neural network to calculate these fractional spin statistics using "entanglement interferometry", and got answers more accurate than the Laughlin wavefunction.
140
David Pfau @davidpfau.com · 22/12/2025
The collective excitations of many electrons behave a lot like particles...but instead of particles with integer spin (bosons) or half-integer spins (fermions) they behave like they have weird fractional spins like 1/3, 1/5, etc. This is the *fractional quantum hall effect*.
120
David Pfau @davidpfau.com · 22/12/2025
Fundamental particles are either fermions or bosons. Fermions, like electrons, make up matter, while bosons, like photons, are force carriers. But in two dimensional materials in a magnetic field, things can start to get weird...
140
David Pfau @davidpfau.com · 13/11/2025
But then a weird thing happened...people found it anyway, and actually started citing it like it was a novel result! I guess I accidentally made the top Google result for "Bregman bias-variance" or something.
160
David Pfau @davidpfau.com · 13/11/2025
Trump's approval rating is now roughly comparable to what it was in mid-2020, at the height of COVID unemployment and George Floyd protests.
070
David Pfau @davidpfau.com · 02/11/2025
The original reply guys.
A page from the Talmud, with commentary from later authors around the original text, and further commentary on the commentary from even later authors around that.
0150
David Pfau @davidpfau.com · 16/10/2025
Meanwhile Commonwealth Fusion Systems is building SPARC, which will be the highest magnetic field tokamak in the world when it comes online in a few years. It is also likely to be the first magnetic confinement device to generate more energy than it consumes. (Image: Hsu and Wurzel 2025)
1191
David Pfau @davidpfau.com · 16/09/2025
Trump's approval rating has been fairly steady, but his favorables have been falling off a cliff. Not really sure what to make of this - people are just happy their 401(k) is going up but increasingly hate Trump himself? (Source: Real Clear Polling)
000
David Pfau @davidpfau.com · 06/09/2025
Can't wait for DARPA to rebrand as WARPA, relocate to London and start releasing critically-acclaimed electronic music.
1271
David Pfau @davidpfau.com · 14/07/2025
I'll be at ICML this week, presenting our paper on Wasserstein Policy Optimization on Tuesday! If you're in Vancouver, come say hi!
1314
David Pfau @davidpfau.com · 05/06/2025
I'm afraid it's not working at the moment, this site is shrinking anyway. And it's shrinking in large part because people find it too much of a monoculture.
110
David Pfau @davidpfau.com · 01/06/2025
I don't often beat up on the NYTimes, because I think the alternatives are mostly worse, but every once in a while they are so egregious my jaw literally drops. Left: NYTimes. Right: The Telegraph (a paper so right-wing it's often called "the Torygraph")
1233
David Pfau @davidpfau.com · 28/05/2025
The war on science in the US is already having an effect on private sector research like AlphaFold. Bears repeating but the private sector builds on top of things created by academic research for the public good. This hurts everyone.
2236103
David Pfau @davidpfau.com · 26/05/2025
I'm enjoying Andor season 2 but every couple of scenes I see something that looks obviously AI-generated and I feel like I'm going insane.
040
David Pfau @davidpfau.com · 02/05/2025
In some sense, there’s nothing in this paper that we couldn’t have done in 2018 (and I wish we had! I’d be famous!) But the inspiration for this paper actually came from the fantastic recent work on Wasserstein QMC by @k-neklyudov.bsky.social and others. Good research should be timeless.
1130
David Pfau @davidpfau.com · 02/05/2025
Tellingly, we have some preliminary evidence that, as the task dimension increases, WPO gets *better* relative to other methods. But it’s surprisingly hard to find truly high dimensional (>100) RL benchmarks. Something for the community to work on!
1100
David Pfau @davidpfau.com · 02/05/2025
This update can be used as a drop-in replacement in almost any actor-critic method in continuous action spaces, and we find that a KL-regularized version of WPO is competitive with other methods on dm_control and a simple tokamak magnetic control task.
181
David Pfau @davidpfau.com · 02/05/2025
Previous work using Wasserstein gradient flows for continuous control arrived at methods similar to SAC which use the reparameterization trick, but we found that there is a simple and general closed form update that works for *any* stochastic policy.
170
David Pfau @davidpfau.com · 02/05/2025
Most RL algorithms for continuous actions fall into two classes. PPO, MPO, etc... work for arbitrary policies but only use scalar value information. DDPG, SAC, TD3, etc… use gradients of the value but are limited to policies that can be represented by reparameterization.
160
David Pfau @davidpfau.com · 02/05/2025
New paper accepted to ICML! We present a novel policy optimization algorithm for continuous control with a simple closed form which generalizes DDPG, SAC etc. to generic stochastic policies: Wasserstein Policy Optimization (WPO).
29410
David Pfau @davidpfau.com · 21/04/2025
Spotted at a pancake restaurant in London. Sign of the times.
0250
David Pfau @davidpfau.com · 08/04/2025
I was basically right there! We were on Lake Salem in Derby.
030
David Pfau @davidpfau.com · 04/04/2025
$NVDA below $100 today. It was about $140 when this photo was taken the day after the inauguration. I think in a generation we'll look back on this as the peak of the generative AI bubble.
113311
David Pfau @davidpfau.com · 25/03/2025
000
David Pfau @davidpfau.com · 19/03/2025
Ok this is getting out of hand...
0121
David Pfau @davidpfau.com · 26/02/2025
050
David Pfau @davidpfau.com · 26/02/2025
2181
David Pfau @davidpfau.com · 26/02/2025
It was masked somewhat by increased immigration from other countries due to the new points-based system, but overall European migration is net negative. People are leaving.
110
David Pfau @davidpfau.com · 20/02/2025
I'm not sure if we can survive another 47 months of this, but at least the honeymoon appears to be over.
4290
David Pfau @davidpfau.com · 03/02/2025
Imagine telling someone in the '90s that "President Trump might shut down The X-Files" is a real actual thing in the year 2025.
0160
David Pfau @davidpfau.com · 02/02/2025
He's already issued more EOs in two weeks than in his first 100 days last term.
040
David Pfau @davidpfau.com · 02/02/2025
"It's a hedge against instability and inflation! It's digital gold!" Gold is up, fwiw.
2724
David Pfau @davidpfau.com · 25/01/2025
I rarely see imagery today that makes it viscerally feel like I'm living in the year 2025, but this definitely looks like something out of dystopian sci-fi.
3886
David Pfau @davidpfau.com · 25/01/2025
290
David Pfau @davidpfau.com · 23/01/2025
Same email sent to the EPA. Seems like a form letter coming straight from the White House.
160
David Pfau @davidpfau.com · 23/01/2025
Looks like DEIAtruth@opm.gov is the email for reporting Illicit Wokeness to the government. Sure would be a shame if the account became unusable because of a flood of joke reports.
5489
David Pfau @davidpfau.com · 18/01/2025
The almost total amnesia about Trump's first term is really insane. Bill Gates was singing Trump's praises the first time too. search.app/VAHFVJee2oWh...
2201
David Pfau @davidpfau.com · 13/01/2025
Jesus Christ, Spotify Discover Weekly, not now.
1210
David Pfau @davidpfau.com · 02/01/2025
We need to return to tech companies having names like American Megatrends LLC.
4353
David Pfau @davidpfau.com · 01/01/2025
In the interest of generating more ML/AI discussion here - here's a really interesting looking paper from Mason Kamb and @suryaganguli.bsky.social on an *analytic* theory of diffusion models. t.co/SYkAAh6k4C
17611
David Pfau @davidpfau.com · 26/12/2024
Watching the MAGA blowback to Trump appointing one Indian-American who supports reforming H1-Bs to a minor advisory role
1141
David Pfau @davidpfau.com · 19/12/2024
My brain is so fried that my first thought when I saw my NeurIPS mug that went through the dishwasher one too many times was "wait, is that AI generated?"
2500
David Pfau @davidpfau.com · 18/12/2024
Saw this graph on the other site, and it did make me wonder - why is there so much hand-wringing about this but not professions that skew conservative? Why are there no op-eds about how there need to be more democrats in the oil and gas industry?
3182
David Pfau @davidpfau.com · 22/11/2024
Maybe the biggest blackpill on climate change is that the most influential thing that anyone can do to fight it is become a high-ranking bureaucrat in the CCP.
3161
David Pfau @davidpfau.com · 11/11/2024
"These guys are weird" was absolutely a winning message.
0121
David Pfau @davidpfau.com · 09/11/2024
This quote from Rorty circulated a lot after Trump won the first time. That time, instead, there was a huge reaction against Trump that went basically the opposite direction. This time, hearing about young men going "your body, my choice", I worry it's actually happening.
080
David Pfau @davidpfau.com · 01/11/2024
This is the real prediction market.
1131
David Pfau @davidpfau.com · 22/08/2024
While excited state energies are harder to calculate than ground states, double excitations - where two electrons are excited at once - are even harder. We show results on these systems that match state-of-the-art calculations at the frontier of what is possible today.
130
David Pfau @davidpfau.com · 22/08/2024
On the carbon dimer - a molecule whose excited states are responsible for the green tint of carbon-rich comets - we match experimental results *five times* more accurately than previous gold-standard calculations.
130
David Pfau @davidpfau.com · 22/08/2024
We couldn’t extend our work with neural networks to excited states because existing methods had several limitations. So we developed a new method that is general, robust, and as accurate as state-of-the-art classical methods as you scale it.
130