Drake meme describing how people shun away from constraining PSF with maximal information across the full focal plane, but rush to do cosmology with maximal information from field-level inference.

a man with a mustache is making a silly face ALT: a man with a mustache is making a silly face

😂😂😂😂😂😂😂

Well we do know our instruments much better than we understand the physics on small scales

Advancing Field-level and Simulation-based Inference for Cosmology Field-level inference has recently emerged as a powerful alternative to traditional summary-statistic approaches in the analysis of cosmological data sets. This technique exploits the full information...

SBI Field Level Inference workshop at Perimeter this June

Hurry up!

events.perimeterinstitute.ca/event/1924/

a man wearing a witch hat is surrounded by orange cones ALT: a man wearing a witch hat is surrounded by orange cones

Working on the problem from two completely different angles is the best part of this project we’ve been hammering for months

Matthijs: pristine, elegant, mathematically flawless

Niko: “okay but imagine the posterior is a traffic cone…”

Somehow… we are both correct 🤣🤣

:red sweater guy:

Ahahaha yes yes

Ah
Right right — it’s his voice yes. I mean he does have a nice voice so idk (not that I would want those self hep things personally but I do get the appeal)

Wut? I don’t follow him so I have no clue. Horrible if true

I only became a fan because of True Detective and The Gentlemen

a close up of a man 's face with a smile on his face ALT: a close up of a man 's face with a smile on his face

Apparently Matthew McConaughey was at Duke today and I missed meeting him because I was stuck in telecons/working from home

This is NOT alright alright alright

I’m really upset

a man with a beard is holding a microphone in front of a crowd . ALT: a man with a beard is holding a microphone in front of a crowd .

Minor pet peeve: describing ΛCDM-based analyses as having the “potential to discover new physics.”

They can reveal tensions or inconsistencies that may motivate extensions to the model — but that’s not the same as discovering new physics.

Cmon my man, tone that down

10/10 recommend. It’s really good quality — I have a few shirts and a hoodie.

Really soft nice cotton, very pleasant to wear

Happy bday, arXiv! You are the best!

Funny coincidence is that I chose to wear one of my arXiv shirts (which are top quality btw! Hoodies as well, highly recommend)

I am a science machine 💪💪💪💪

4th paper fresh on the @arxiv.bsky.social this year 🥳🥳🥳

arxiv.org/pdf/2603.09973

a man says it 's very refreshing while wearing a jacket ALT: a man says it 's very refreshing while wearing a jacket

I was in a clusters telecon today (not my usual working group) and am so beyond impressed by the quality of plots people were showing. I find it so refreshing to see there are people out there who care about communicating their research

Figure 1

Figure 2

Figure 3

Figure 4

Impact and interplay of $Λ$CDM analysis choices for LSST cosmic shear. N. C. Robertson et. al. https://arxiv.org/abs/2602.21863

Will do!

It’s still cold here but expecting temps in the 80s this week and in here for it

Starting the day shearing dead weeds in the forest and cutting through mountain mint — that sharp minty scent mixed with damp, rotting leaves — pure peace for the soul 🌿

And catching up with the wonderful folks over at @DukeLemurCenter is just 👌chefs kiss

Also Derivkit assumes you have your model and covariance and if you have that (which you obviously do), you can go wild with forecasts (just choose your derivative method wisely — consult workflows page in the docs)

You are welcome! If you need any clarification please do not hesitate to contact me.

I’d start with either example pages (forecasting) or workflows page (that one has a FAQ and some decision guidance)

That’s one hella parameter space!

Btw if you wanna go down the quicker route do diagnostics (Fisher/DALI), you can do so with derivkit docs.derivkit.org/main/

(Plug but I’m not ashamed 😂)

Two more works on the @arxiv 🥳🥳🥳

arxiv.org/abs/2602.21863

arxiv.org/abs/2602.21280

Here docs.derivkit.org/main/index.h...

Btw @rjme.bsky.social did you see the work Matthijs, a CMB colleague and I put up on arXiv the other day? Perhaps your students might be interested in trying it out

Didnt*

I also don’t know what’s up with autocorrect

I don’t know that!!! I do indeed smile a lot. Here’s me smiling with one helluva guy (photo taken few years ago)

Aaaaaand it’s officially on @arxiv.bsky.social 🥳🥳

arxiv.org/abs/2602.08078

Wanna do Fisher, XY Fisher or DALI with a chosen derivative method backend in a simple sleek Python way? Look no further, my friend. We got you!

Simple examples here

docs.derivkit.org/main/example...

Ok, aside that this was ten years ago (wthhhhell 😱) and I now feel ancient, it is actually nice to be reminded of good times during masters in Göttingen: Thorsten was my lecturer in cosmology and string theory courses and was super encouraging. Made me fall in love with cosmo 🤩

a cartoon of a tiger and a snake with the words " i have no time for that sort of nonsense " ALT: a cartoon of a tiger and a snake with the words " i have no time for that sort of nonsense "

Rest? 🤣🤣

Hahah oh no, we’re on a winter break.

We’ve been working every single day every day on the thing you are pushing us for 🤣

Actually — the first burek in the states since I moved 😱

Found the ingredients at Whole Foods

Also a selfie just because