Tootfinder

Orwell wrote
Nineteen Eighty-Four
not as a prophecy
but as an extrapolation and a warning.
As he explained:
“I do not believe that the kind of society I describe necessarily will arrive.
But I believe, allowing, of course, for the fact that the book is a satire,
that something resembling it could arrive.”

1984 revisited: George Orwell would be relieved at how we’ve done
It’s common to suggest the great dystopian novel was full of warnings that have largely come true. But the evidence shows otherwise

8 Things Veganism is NOT #AnimalRights

8 Things Veganism is NOT
Jordi Casamitjana, author of the book “Ethical Vegan”, discusses eight things that veganism is not, but many people erroneously believe it is We live in difficult times. For a few years now…

Exploiting ID-Text Complementarity via Ensembling for Sequential Recommendation
Liam Collins, Bhuvesh Kumar, Clark Mingxuan Ju, Tong Zhao, Donald Loveland, Leonardo Neves, Neil Shah
https://arxiv.org/abs/2512.17820 https://arxiv.org/pdf/2512.17820 https://arxiv.org/html/2512.17820
arXiv:2512.17820v1 Announce Type: new
Abstract: Modern Sequential Recommendation (SR) models commonly utilize modality features to represent items, motivated in large part by recent advancements in language and vision modeling. To do so, several works completely replace ID embeddings with modality embeddings, claiming that modality embeddings render ID embeddings unnecessary because they can match or even exceed ID embedding performance. On the other hand, many works jointly utilize ID and modality features, but posit that complex fusion strategies, such as multi-stage training and/or intricate alignment architectures, are necessary for this joint utilization. However, underlying both these lines of work is a lack of understanding of the complementarity of ID and modality features. In this work, we address this gap by studying the complementarity of ID- and text-based SR models. We show that these models do learn complementary signals, meaning that either should provide performance gain when used properly alongside the other. Motivated by this, we propose a new SR method that preserves ID-text complementarity through independent model training, then harnesses it through a simple ensembling strategy. Despite this method's simplicity, we show it outperforms several competitive SR baselines, implying that both ID and text features are necessary to achieve state-of-the-art SR performance but complex fusion architectures are not.
toXiv_bot_toot

Eieiei, ich bin vollkommen baff. Ich war gerade mit meinem Herzmenschen in Darmstadt bei Habibi in der Nieder-Ramstädter-Straße essen. Noch bevor wir an der Speißekarte waren gab es einen Falafel und ein Becherchen vegane Linsensuppe zum Probieren. Mjam! Zwei vegane Hauptspeisen bestellt und als sie fertig waren, bot uns der Besitzer noch Kuchen als Nachtisch aufs Haus an, der war allerdings mit Joghurt, weswegen wir ihn freundlich abgelehnt haben. Nach zehn Minuten bringt er uns zwei frisch gebackene Kekse mit Dattel-Kardamom-Füllung! Wie verrückt! Wir wurden nach Strich und Faden verwöhnt.
Das gab natürlich ein ordentliches Trinkgeld. Ob wir Lust hätten, noch Falafel für unterwegs mitzunehmen, fragte er. Das nahmen wir natürlich dankend an. Und jetzt zu Hause machen wir die Box auf: NEUN Stück! Der helle kulinarische Wahnsinn.
Nicht nur wegen der vielen Extras können wir den Laden nur wärmstens empfehlen! Das war echt unglaublich lecker 🤤
#Darmstadt
#Habibi
#HabibiDA
#SyrischerImbiss
#Falafel
#vegan