abbassix

For those reading the paper: the positives of a good-view JEPA are described as “two observations of one scene,” and yet the paper also states that “A good-view JEPA's positives are (g_m(z, ν), g_m(z′, ν′)) with (z, z′) ∼ π and independent nuisances.” My question is: why doesn't “two observations of one scene” mean that both observations share the same underlying world state (z) and differ only in their modality nuisances (ν and ν′)? Also, I don't understand what “a symmetric positive-pair distribution (z, z′) ∼ π_m” (or more explicitly, a “world kernel”) means.

By the way, the server is down until Monday, when we'll be able to reach the office.

Unless, of course, only the first attempts count toward the leaderboard.

The repeated images might be due to a ranking algorithm that samples random images. Since not all crops are solvable, images where many people achieved a decent score end up ranking higher.

There are thousands of images and thousands of different crops from the images. Some are not solvable, but on average you can solve most of them!

Given two crops of the same image, determine their position relative to each other.

"Flying Machines Which Do Not Fly" is an editorial published in the New York Times on October 9, 1903. The article incorrectly predicted it would take one million to ten million years for humanity to develop an operating flying machine. It was written in response to Samuel Langley's failed airplane experiment two days prior. Sixty-nine days after the article's publication, American brothers Orville and Wilbur Wright successfully achieved the first heavier-than-air flight on December 17, 1903, at Kitty Hawk, North Carolina.

In her essay Economia di guerra e inquadramento della società in URSS (The War Economy and Social Control in the USSR), Rebecca Manley writes that initially the German advance in Russia was welcomed by the local population and threatened the legitimacy of the reigning power, while party leaders, officials, and factory managers often jumped ship, with the abandoned populace openly cursing them [A. Aglan and R. Frank, La guerra-mondo 1937-1947 (The World War 1937-1947), Einaudi, 2015-16]. Her account goes on to describe the government's gradual reassertion of control: The cost in human lives was increasing and shortages were continuing to worsen [...] it was only with the victory of Stalingrad that Stalin and the Soviet system recovered a central role in war propaganda. Overall, according to this author, the war years saw a surge of popular patriotism, even among those who had reason to distrust the State. The authorities in Moscow maintained order on the home front, something that those of the Tsarist Empire had failed to do.

The disorientation of non-Russian nationalities In the Russian countryside, and particularly in the three Baltic countries, Ukraine, the Caucasus, and Belarus, many had anxiously waited for the Germans as saviours and liberators from the Russian yoke [István Deák, Europe on Trial, Hachette UK, 2015]. However, the peasants soon felt deceived by those in whom they had initially trusted, because the Germans did not dismantle the collective farms (the kolkhoz) into which their lands had been forcibly incorporated. The Germans conserved the kolkhoz, because they were afraid production would fall otherwise, and, on the contrary, increased forced requisitions. The Russian government did form assault groups, recruited among the detainees in the forced labour camps, but, in the main, these were anti-Bolshevik Russian nationalists, motivated by patriotism. In contrast, prisoners of subjugated nationalities, such as the Poles, Ukrainians, and Balts, often escaped to join the Germans. On the whole, releasing anti-Stalinist political prisoners to enrol them in the army was forbidden; indeed, as we have already seen [Internationalism, April 2026], standing orders were to shoot all prisoners in the event of a forced retreat due to the German advance.

In assessing the present-day irruption of artificial intelligence, especially so-called generative AI based on highly effective linguistic-probabilistic systems, the starting point is to recognise that what is at stake is the development of the productive forces. The next step is to relate the critical aspects of AI – so hotly debated in the struggle between capitals who have a stake in it – not primarily to the intrinsic character of the technology itself, but to the fundamental contradiction between the development of the productive forces and capitalist relations of production.

First, for Engels, the brain is a product of labour. It is in the dialectical relationship of mutual action and reaction with labour – made possible by the articulation of the hand freed by man's upright posture, the result of "hundreds of thousands of years" of natural selection – that the brain evolved to perform the most complex functions and develop self-awareness. In turn, labour is an expression of the social relations at the basis of specific socio-economic formations across successive modes of production. Therefore, the brain is a social product, mediated by the hand that wields the tool, by labour that puts it in relation with the brains of other humans, and by the language made necessary by those relations.

Second, this relationship of upright posture-hand-language-brain, and the role played by labour in humanisation, is neglected by current conceptions and is also unrecognised by Darwinian evolutionists themselves. This is because the division into classes and the separation between manual labour and intellectual labour – in the latter's connection with the ruling class and its apparatus (schools, the Churches, etc.) – has led to the prevalence of the idealist conception that gives prominence to the brain and its representations.

Third, humanity has applied science to nature, and nature has thereby been modified through science, technology, and labour. But it has been difficult to go beyond the initial and immediate effects of this transformation, barely taking into consideration its secondary and long-term effects.

Fourth, and even more importantly, science has not been applied to the social consequences of labour and to its implications, from the division of labour to class struggle, thus depriving humanity of a vision of the wider, long-term effects of productive activity. Bourgeois economic science, i.e., classical political economy, has been unable to go any further. In order to regulate human labour activity according to a truly conscious plan, knowledge is not enough; an overturning of the mode of production and social order is necessary.

Fifth, the individual capitalist is concerned only with immediate profit, not with what will happen to the commodity or the buyer afterwards. The same applies to what will happen to nature and – we might add – to human nature itself.

More wars; there could be even a world war! All caused by imperialist competition between powers: the US, China, the EU, India, Brazil, Russia, etc. More nationalism (even with continental flavor like European nationalism), more militarism, more revolts, and possibly internationalist revolutions to unite the global working class against all wars and nationalisms!

I see these wave-like patterns on this lake. Are these aliasing artifacts?

"It [Anthropic] said DeepSeek's operation involved over 150,000 exchanges". In my humble opinion, a mere 150k exchange for an LLM could only be a benchmarking and not a distillation! I think the US companies should accept that after decades they have rivals surpassing them, just like they did Europeans almost a century ago.

Gulags were the worst kind of exploitation; that is why we always were anti-Stalinist and always identified the USSR as 'state capitalism'! The fact that a state 'claims' socialism does not mean it 'is' socialist, just as North Korea claiming 'democratic' people's republic does not mean it is 'democratic'. Clear?

Our heros fought and died for 8-hour working days and when we stop fighting, we go back to 12-hour working days! That is why you need to search for theinternationalism.org

We are close to treating PTSD of people who we sent to massacre other people, and we justified their massacre as heroic act. Great achievement for humanity!

Should we believe that intelligence agencies like CIA or FBI doesn't know her/him?

I think if you support Netanyahu, you are not in a position to condemn these atrocities. The problem is that Iranian pro-democracy opposition is demolished by far-right sometimes neo-Nazi monarchists!

I bought the latest Intel MacBook Air and after few months they released M1 MacBook Air. Then they release an update, I did as I should, and it broke my laptop! I lost all my data! They fixed it for me in a couple of weeks without providing me a substitute! And they were arrogant enough to behave like they did a favor to not charge me for the repair! It was 2020! My MacBook Air never became like before, and it died a couple of times in a year, which it recovered by itself. And some months ago it died, and it didn't recover! I lost lots of data I spent thousands of hours and I trusted that I can have them even if my laptop dies. Now, I have no way to recover my data!

We still do not have a standardized way to represent Machine Learning concepts. For example in vision model, I see lots of papers confused about the "skip connections" and "residual connection" and when they concatenate channels they call them "residual connection" while it shows that they haven't understood why we call them "residual" in the first place. In my humble opinion, each conference, and better be a confederation of conferences, work together to provide a glossary, a technical guideline, and also a special machine translation tool, to correct a non-clear-with-lots-of-grammatical-error-English like mine!

In his second term, Donald Trump launched a sweeping overhaul of U.S. economic and political strategy, marked by historic tariff hikes, aggressive fossil fuel expansion, and a dramatic shift toward cryptocurrency deregulation. His executive orders dismantle climate commitments, target federal spending, and centralize power through new institutions like the Department of Government Efficiency. With the appointment of an AI & Crypto Czar and rejection of central bank digital currencies, Trump pushes for private control over digital assets, raising concerns about deregulation and monetary instability. These moves signal a sharp break from past policies, reshaping inter-imperialist relations and global economic dynamics.

BYD's leap forward is impressive, comparable to that of Ford Motors after the First World War, when with the Model T, introduced in 1908, it revolutionised the auto industry and ushered in the motorisation of the 20th century. In just five years, from 1920 to 1924, Ford's sales quadrupled from 460,000 to 1.9 million. Tesla Motors made a comparable leap forward in 2020-24, more than tripling its sales from 500,000 to 1.7 million.

The «Communes» campaign has posed a fundamental problem for the Marxist interpretation of the development from a semi-feudal economy to a capitalist economy. First of all, we need to grasp clearly that in China, as in every other country with a feudal or semi-feudal economy, the historical problem is that of creating a capitalist economy, since it is impossible to skip a whole stage of economic development. Only when the material bases are created – and the creation of these bases is the task of capitalism, in this case of state capitalism – will Chinese society, and for it the new revolutionary class created by industrialisation, i.e. the proletariat, be mature enough to raise the problem of the socialist management of its economy.

Paralysed man of spinal cord injury, once considered incurable, walked again in Shanghai’s Zhongshan Hospital. The “three in one” brain-spine interface uses electronic implants wirelessly transmit patient’s thoughts to his legs and feet via newly-created “neural bridge” on his spine.

Professor Jia’s team spent nearly three years developing a lightweight AI model to decode brain signals in real time and simulate right electrical stimulation to achieve muscle movement.

This is the first time such a complete system has been achieved worldwide. Early this year, the research officially entered the clinical concept validation phase.

Mathematicians recently proved a central component of the Langlands program, an ambitious effort to develop a “grand unified theory” of mathematics. The monumental proof of the geometric Langlands conjecture totaled more than 800 pages and marked the culmination of 30 years of work by nine mathematicians.

Hear from the Microsoft team behind the recent breakthrough in physics and quantum computing demonstrated by the new Majorana 1 chip, engineered from an entirely new material that has the potential to scale to millions of qubits on a single chip. Find out what is possible...