Lo de MariCarmen va a ser el sacrificio del eclipse de la política Española. Como gane Pedro, se cepillan a frijolito y tenemos a Ayuso como nueva integrante de la mano de Dios.

a la cama

La Mari Carmen va a tener hasta su ninot indultat en las fallas de Valencia. Increíble este evento canónico, va a ser el eclipse de Berserk de la derecha en este país.

Please China invade us we have plenty of rice in Valencia.

Sinceramente espero que la calentada de perro le salga cara. La sanidad está colapsada, la educación también y hecha para el más retard (dando responsabilidad total al maestro o profesor) y la vivienda lo mismo. Como Carmena, mucha izquierda pero sólo para los de Malasaña.

If you stop messaging your friends after months of initiating conversation and they never write first themselves. Are they friends?

Ok let's see. Indians smell like shit.

Un puto moro sorprendido porque los blancos no abusan directamente de la tipa. Luego las tipas sorprendidas cuando un moro les viola y les corta el cuellom

2 years of testing FSD Supervised all over Spain Across 460,000 km, we have seen it all

Es increíble que muchos jóvenes españoles se tienen que basicamente alejarse de sus familias y amigos para trabajar. Y ahora toda la puta panchitada se está juntando en España y destrozando aún más las perspectivas de esos jóvenes.

A lot of hype around OpenAI's Astra model here on my timeline today. Apparently, this goes back to a new article from The Information, which said Astra is a "recurrent depth or looped transformer". It's always interesting to read about new or different approaches (including rumors about what the closed labs may be up to), but let's debunk this a bit. About 2 months ago, I shared the architecture details of Nanbeige, for example, where "Nanbeige4.2-3B is pretrained from scratch on 28T tokens with a Looped Transformer that reuses the layer stack to increase capacity without adding parameters." Yes, that's it. The looped transformer idea is just reusing layers in the transformer block. In the case of Nanbeige, the main idea is to reuse the same 22-layer stack (=transformer block) twice instead of once. So, effectively it extends the 22-layer architecture to 44 layers, but without duplicating the weights. In simple terms, this roughly doubles the size of the model (if we ignore the embedding and output layers for a second). But instead of requiring 2x the storage and RAM to host this model, it stays at the same size since we reuse the components. However, it's almost 2x as expensive in terms of compute, because we run the embedded text through almost 2x as many layers. Why? In the Nanbeige 4.2 technical report, the researchers found that two passes gave the best trade-off and retained about 75% of the token efficiency of a standard architecture. (More passes gave barely any gains but made the training much slower and much more expensive.) While, as far as I know, Nanbeige 4.2 is the first notable open-weight model that adopted this approach, the idea goes back to the NeurIPS paper "Mixture-of-recursions: Learning dynamic recursive depths for adaptive token-level computation". Actually, this paper proposes a mechanism that is a bit more sophisticated by adding a learned router that determines whether each token receives one, two, or more passes. So, easy tokens can exit early while harder tokens receive additional computation. In sum, Astra may be a really good model, but this shouldn't be about this "looped transformer aspect," which is just a tiny architectural tweak. Also, the statement "the new technique works in a way that obscures some or all of the AI's reasoning, otherwise known as 'chain-of-thought'" is not necessarily true with respect to the looped transformer method. It's possible that The Information journalist refers to some other technique or misunderstood the looped transformer method. Reusing layers does not by itself suppress visible chain of thought. It adds computation in hidden states before the next token is emitted, just as ordinary transformer layers do. But based on the information we have, the only plausible interpretation here is that if a model uses more of these recurrent passes, it may need to generate fewer intermediate reasoning tokens. So then more of its computation happens in latent activations that cannot be read as text. But we would get the same effect if we were scaling up the model size, like GPT 5.6 Luna -> GPT 5.6 Sol.

Se folla a una pancha 100% porque hoy en día es inaguantable.

Con la panchificación de toda la hostelería es increíble lo al tanto que tienes que estar con las vueltas. A veces pago con tarjeta sólo para no tener que contar el cambio.

Goldmine for Software Engineers! 📌 If you're learning System Design and/or AI Engineering, then bookmark this repo right now. github.com/systemdesign42/sy…

Why are so many women sooo angry? They won. They got everything they wanted. Men have never been doing this badly, and women have never been doing this well. They make more, they have the majority of jobs, they are the majority of college graduates, (UCLA undergrad is 60% female), they make more money, have much lower rates of suicide, addiction, and alcoholism, and live longer. They won: they took over all men-only spaces, they took over the Boy Scouts, they ended men’s bathrooms, they can fly fighter planes and be in combat. Marriage is optional, divorce is no-fault, they win 99% of the custody battles, they can control fertility at will, and they can’t get drafted. We are all living in the female future they promised.

La proyección: