Adrien Corenflos
@adriencorenflos.bsky.social
Assistant Professor at the university of Warwick.
I compute integrals for a living.
https://adriencorenflos.github.io/
I compute integrals for a living.
https://adriencorenflos.github.io/
I like big words too, they take a lot of space in my reader's mind.
November 14, 2025 at 7:44 PM
I like big words too, they take a lot of space in my reader's mind.
Say you're lazy and you use an LLM to do your review. Surely, when the LLM returns a *this* you'll have a double take "wow, that's a bit long!"
November 14, 2025 at 7:33 PM
Say you're lazy and you use an LLM to do your review. Surely, when the LLM returns a *this* you'll have a double take "wow, that's a bit long!"
I saw it. It's state-of-the-art trolling. Too bad there's no track for such valuable contributions.
November 14, 2025 at 7:12 PM
I saw it. It's state-of-the-art trolling. Too bad there's no track for such valuable contributions.
Forgot a slash there
arxiv.org/abs/2511.06351
arxiv.org/abs/2511.06351
arxiv.org
November 11, 2025 at 10:53 AM
Forgot a slash there
arxiv.org/abs/2511.06351
arxiv.org/abs/2511.06351
Oh that's pretty fun! I'll give it a read in both languages!
November 2, 2025 at 5:15 PM
Oh that's pretty fun! I'll give it a read in both languages!
www.lbpedlar.com
It's their weekend-special pastry. They have some limited pieces that vary every Saturday. It was 5£ mind you so I do expect it was a non-negligible amount of work :)
It's their weekend-special pastry. They have some limited pieces that vary every Saturday. It was 5£ mind you so I do expect it was a non-negligible amount of work :)
Home | Little Bread Pedlar
London-based bakery providing artisan bread & buns, plus sweet & savoury pastries to take away.
www.lbpedlar.com
November 1, 2025 at 10:29 AM
www.lbpedlar.com
It's their weekend-special pastry. They have some limited pieces that vary every Saturday. It was 5£ mind you so I do expect it was a non-negligible amount of work :)
It's their weekend-special pastry. They have some limited pieces that vary every Saturday. It was 5£ mind you so I do expect it was a non-negligible amount of work :)
I think I figured out, it's a mix between standard puff (inside) and a casing of inverted puff. I'll ask them tomorrow.
November 1, 2025 at 10:25 AM
I think I figured out, it's a mix between standard puff (inside) and a casing of inverted puff. I'll ask them tomorrow.
Honestly no idea how they even made that one. The inner was fairly traditional lamination, but that shell 🤷
November 1, 2025 at 10:21 AM
Honestly no idea how they even made that one. The inner was fairly traditional lamination, but that shell 🤷
"historically, generative models have"
October 31, 2025 at 9:06 AM
"historically, generative models have"
Reposted by Adrien Corenflos
Surprise 1: it is possible!
Surprise 2: the algorithms are quite simple!
Surprise 3: it is beautiful (this one is not a surprise)
For a good overview and some cool results on correlated sampling, see, e.g., arxiv.org/abs/1612.01041
(Now, are you happy, tiny part of my brain I can't ignore?)
Surprise 2: the algorithms are quite simple!
Surprise 3: it is beautiful (this one is not a surprise)
For a good overview and some cool results on correlated sampling, see, e.g., arxiv.org/abs/1612.01041
(Now, are you happy, tiny part of my brain I can't ignore?)
Optimality of Correlated Sampling Strategies
In the "correlated sampling" problem, two players are given probability distributions $P$ and $Q$, respectively, over the same finite set, with access to shared randomness. Without any communication, ...
arxiv.org
October 27, 2025 at 11:16 AM
Surprise 1: it is possible!
Surprise 2: the algorithms are quite simple!
Surprise 3: it is beautiful (this one is not a surprise)
For a good overview and some cool results on correlated sampling, see, e.g., arxiv.org/abs/1612.01041
(Now, are you happy, tiny part of my brain I can't ignore?)
Surprise 2: the algorithms are quite simple!
Surprise 3: it is beautiful (this one is not a surprise)
For a good overview and some cool results on correlated sampling, see, e.g., arxiv.org/abs/1612.01041
(Now, are you happy, tiny part of my brain I can't ignore?)
Assumption A: it rains every day, Assumption B: it rains sometimes. E: it rains now. E is clearly more likely under A so we should favour A all the time. This is of course sophistic because we have *strong* evidence that A is quite unlikely.
October 24, 2025 at 10:20 PM
Assumption A: it rains every day, Assumption B: it rains sometimes. E: it rains now. E is clearly more likely under A so we should favour A all the time. This is of course sophistic because we have *strong* evidence that A is quite unlikely.
Honestly from my perspective I have, so that's that. Talking clearly, I just can't accept an argument based on evidence isolated from prior likelihood. E.g. your reasoning has the same feel to me as (...)
October 24, 2025 at 10:20 PM
Honestly from my perspective I have, so that's that. Talking clearly, I just can't accept an argument based on evidence isolated from prior likelihood. E.g. your reasoning has the same feel to me as (...)
Agree to disagree, I'd rather not waste time in nitpicking. You don't seem very happy to revise your assumptions and I'm probably not either! Interesting read either way and thanks for sharing
October 24, 2025 at 9:55 PM
Agree to disagree, I'd rather not waste time in nitpicking. You don't seem very happy to revise your assumptions and I'm probably not either! Interesting read either way and thanks for sharing
Right. Pardon my directness, but I think there's quite a bit of difference between something being directly compatible with the most successful model of reality we have, and your supposed hallucinations.
October 24, 2025 at 9:50 PM
Right. Pardon my directness, but I think there's quite a bit of difference between something being directly compatible with the most successful model of reality we have, and your supposed hallucinations.
I think I am? I'm specifically stating you are not using the most specific evidence you have.
October 24, 2025 at 9:46 PM
I think I am? I'm specifically stating you are not using the most specific evidence you have.
There's quite a bit more than in favour of god existing though: it's compatible with pretty much everything we know to be true.
October 24, 2025 at 8:13 PM
There's quite a bit more than in favour of god existing though: it's compatible with pretty much everything we know to be true.
While it's very unlikely for a single burglar to crack their safe, it's not unlikely for one to. I have 0 chance to win the lottery, but someone's gonna win it. Arguably this is by design, and nothing prevents you from pushing back the hand of god to "designing the multiverse such that it works so".
October 24, 2025 at 5:36 PM
While it's very unlikely for a single burglar to crack their safe, it's not unlikely for one to. I have 0 chance to win the lottery, but someone's gonna win it. Arguably this is by design, and nothing prevents you from pushing back the hand of god to "designing the multiverse such that it works so".
To be precise, I think this is the most specific evidence we have.
Now, in terms of your argument in itself, independent of the disagreement on the evidence ratio missing prior, I think your safe/burglar analogy doesn't hold: it's not one safe one burglar, it's many safes many burglars.
Now, in terms of your argument in itself, independent of the disagreement on the evidence ratio missing prior, I think your safe/burglar analogy doesn't hold: it's not one safe one burglar, it's many safes many burglars.
October 24, 2025 at 5:36 PM
To be precise, I think this is the most specific evidence we have.
Now, in terms of your argument in itself, independent of the disagreement on the evidence ratio missing prior, I think your safe/burglar analogy doesn't hold: it's not one safe one burglar, it's many safes many burglars.
Now, in terms of your argument in itself, independent of the disagreement on the evidence ratio missing prior, I think your safe/burglar analogy doesn't hold: it's not one safe one burglar, it's many safes many burglars.
I don't understand the logic: there is *some* prior evidence in favour of the multiverse because it emerges naturally from known (albeit speculative) physics. There is no known physics from which one deduces the existence of a god. It seems more Bayesian to me to hence conclude the exact opposite.
October 24, 2025 at 5:25 PM
I don't understand the logic: there is *some* prior evidence in favour of the multiverse because it emerges naturally from known (albeit speculative) physics. There is no known physics from which one deduces the existence of a god. It seems more Bayesian to me to hence conclude the exact opposite.
Ack: this example was suggested by P.E. Jacob.
October 20, 2025 at 4:20 PM
Ack: this example was suggested by P.E. Jacob.