#generalisation
if i see another person make an unfair generalisation I'm going to stand up and yell
November 21, 2026 at 12:47 PM
Major #neurogam update: now supports the analysis of 1D time courses, 2D temporal generalisation, and 3D sensor x time (M/EEG) data. CRAN version coming soon!

📦 R package: lnalborczyk.github.io/neurogam/
📄 Updated preprint: www.biorxiv.org/content/10.1...

#rstats #brms #EEG
November 20, 2026 at 5:12 PM
That’s a sweeping generalisation 😂
November 21, 2026 at 7:21 PM
Avant de chercher à l'étendre à d'autres crimes (parce que c'est comme ça que la loi pénale fonctionne dans ce pays : d'abord l'exception, puis la généralisation).
November 21, 2026 at 7:24 AM
It's almost as if bigotry and being a massive wrongun are closely linked character traits.
I'm sure some would say that's a massive generalisation...but if it quacks like a duck...
extremely bizarre clip where Graham Linehan claims that convicted rapist and paedophile Stuart Hall was an innocent man merely “swept up” in Operation Yewtree
November 18, 2026 at 1:10 PM
Sweeping generalisation from this employer about young people. I wouldn't want to work for her
#R4Today
November 17, 2026 at 7:42 AM
Comme l'impression que les blocus de lycée s'épuisent.
Difficile de dire si c'est du fait de la généralisation du distanciel depuis 8 jours.
November 20, 2026 at 4:27 PM
“Why do people think AI caused [disaster]?”

You spent untold amounts of money and effort trumpeting the rise of the Hasty Generalisation Algorithm — including by foretelling it causing disasters — and then have the gall to object when people short-circuit-inferred it as the cause of disasters?
November 18, 2026 at 6:17 AM
It's almost as if bigotry and being a massive wrongun are closely linked character traits.
I'm sure some would say that's a massive generalisation...but if it quacks like a duck...
November 18, 2026 at 1:10 PM
This implies that there are young people sitting at home, on benefits & living the life of Riley.
Nothing could be further from the truth.
An arrogant & ignorant generalisation.
Not all people are equal.

#TaxWealthNotWork
November 17, 2026 at 10:16 AM
Harvard researchers uncovered a link between model size and "grokking" in neural networks, demonstrating that generalization versus memorization in small Transformers depends on distinct timescales. https://arxiv.org/abs/2605.09724
How Large Must a Model Be to Grok? A competition between memorisation and generalisation timescales
ArXiv link for How Large Must a Model Be to Grok? A competition between memorisation and generalisation timescales
arxiv.org
November 18, 2026 at 8:50 AM
This one's a doozy...
November 19, 2026 at 7:49 PM
The crass generalisation of an imagined problem group is grim.

Immigrants, young people, disabled people, low earners etc.

The logical conclusion to this train of thought doesn't bear thinking about.
November 17, 2026 at 10:58 AM
C’est parce que je vois et la dimension politique et son exploitation politicienne que cette vidéo est salutaire. On peut vivre en mode binaire ou sortir de la pseudo pensée radicale qui ne pense plus. Refuser la généralisation ce n’est pas nier les problèmes structurels. C’est éviter le simplisme.
November 20, 2026 at 6:20 PM
That’s quite a generalisation. How many ‘atheists’ are you including?
November 20, 2026 at 8:50 AM
WOW!!

That’s a BIG generalisation to make!!

You don’t know who would like those movies. No one does!
November 17, 2026 at 2:30 AM
J’aimerais sincèrement comprendre ce qu’apporte une généralisation qui est factuellement fausse. Le fait que des hommes agressent des femmes en toute impunité n’implique pas que tous les hommes soient agresseurs. Parler ainsi c’est essentialiser les VSS et invisibiliser leur caractère systémique 1/
November 21, 2026 at 8:16 PM
Ce slogan repose sur le principe de précaution rationnelle, c'est une généralisation statistiques, pas universelle. Pas tous les hommes mais toujours un homme, pas tous les flics mais tjr un flics, sans parler des autres facteurs de banalisation/minimisation des actes sans pr autant en être l'auteur
November 21, 2026 at 9:50 PM
Just qwen and a few others messing around with unsloth.

But I didn't even know thinking/coding fitted on my gpu... let alone ran at more than 1 token per second.

If it's competent (small niche model, I don't expect generalisation) it's *usable*

Amazing.
November 19, 2026 at 12:53 PM
Negative experiences being

1. Generalisation of my people’s skin tone
2. Beauty

Etc..
November 19, 2026 at 6:28 PM
Le groupe « socialistes et apparentés » a fait voter à l'Assemblée Nationale ce matin la loi sur la généralisation de l’expérimentation "Territoires Zéro Chômeur de Longue Durée”. Cela permet de combattre le chômage de longue durée, en s’adaptant à la réalité de chaque territoire.
November 19, 2026 at 9:56 AM
Il n'y a qu'une opposition massive à l'imposition des IA dans nos vies, la conviction profonde que la généralisation des IA relève de la dystopie à tous les niveaux qui nous permettra de maintenir (ou de reconstruire) la capacité de vivre sans les IA
November 15, 2026 at 5:26 PM
I don't disagree. The way these COBOL stories are often presented implies these entities have no choice. They're also presented as some sort of world-wide generalisation that banks are still using COBOL.

I don't believe that for a minute (other than perhaps in the US, where everything is stupid). 🤡
November 18, 2026 at 2:51 AM
Very pleased to share a new preprint! 🎉

Work by Phoenix Pavlik (an undergrad student in our group) shows that people generalise learning rates from prior experience

Exposure to a stable environment makes people treat future environments as more stable

osf.io/preprints/ps...
How Past Experiences Shape New Beliefs: Generalisation and Adaptation in Human Reinforcement Learning
Humans are adept at learning under uncertainty. However, the way in which we learn must adapt to different environments, and we must determine when to generalise learning strategies across different c...
osf.io
November 17, 2026 at 4:07 PM