📌 TOPINDIATOURS Update ai: US scientists make Josephson junction using only one su

Researchers at the University of Buffalo could add a ‘superconducting’ twist to magnetic hard drives and random access memories (RAMs) after succeeding in building a Josephson junction using only one superconductor instead of two. The achievement could unlock new, simpler, and more flexible quantum designs in the future. 

A Josephson junction is the foundational building block of quantum computers. Built on the principle of superconductivity, the device consists of two superconductor layers separated by a thin barrier. Under these conditions, the superconductive nature flows into the barrier, synchronizing their behavior. 

Researchers at the University of Buffalo teamed up with those in Spain, France, and China to demonstrate that a Josephson junction could be built with just one superconductor layer instead of two and in the process showed how commonly available iron could also be used for quantum applications. 

What did the scientists do? 

At the Autonomous University of Madrid, the researchers conducted experiments to fabricate a Josephson junction with a superconducting vanadium electrode on one side and an iron electrode on the other, with magnesium oxide as the barrier. 

While superconductivity allows electrons to flow without energy loss, the flow is not continuous. Much like water flowing through a faucet appears smooth but is made up of individual droplets, continuous current is also made up of individual electrons that flow in tiny bursts. 

“These small, unavoidable fluctuations in electron flow are called noise, and by listening to them we can learn how charge moves through a material,” said Jong Han, PhD, professor in the UB Department of Physics, who was involved in the work. 

By analyzing the noise, the researchers measured the flow of electrons in iron and found it to be much like that in a Josephson junction. 

“A typical Josephson junction with two superconductors is like two army battalions marching in step along opposite banks of a river. In our experiment, there was only one battalion — yet it’s as if its marching caused citizens on the other side to form a militia and begin marching to the beat of a different drum,” explained Igor Žutić, professor in the Department of Physics at the University at Buffalo. 

Why is this important? 

This behavior was unexpected because iron is a ferromagnetic substance, which is quite the opposite of a superconductor. While spins of electron pairs in superconductors are in opposite directions, those in ferromagnets are in only one direction.

Somehow, iron was able to make superconducting electron pairs, even though its spins were in the same direction. The research team does not even have a theory to explain this behavior, but is excited about its potential applications. 

“The problem with conventional quantum computers is that even small environmental changes can throw off the spin of their electrons,” added Zutic. “We want to find a way to lock an electron’s spin into place, essentially, and same-spin pairing could hold some answers.”

Another added advantage of this research is that, in the future, Josephson junctions could be made from ordinary materials, such as iron and magnesium oxide. Both of these materials are commonly used in hard drives and RAMs available today. 

“We have added a superconducting twist to commercially viable devices,” concluded Žutić. 

The research findings were published in the journal Nature Communications.

🔗 Sumber: interestingengineering.com

📌 TOPINDIATOURS Eksklusif ai: Adobe Research Unlocking Long-Term Memory in Video W

Video world models, which predict future frames conditioned on actions, hold immense promise for artificial intelligence, enabling agents to plan and reason in dynamic environments. Recent advancements, particularly with video diffusion models, have shown impressive capabilities in generating realistic future sequences. However, a significant bottleneck remains: maintaining long-term memory. Current models struggle to remember events and states from far in the past due to the high computational cost associated with processing extended sequences using traditional attention layers. This limits their ability to perform complex tasks requiring sustained understanding of a scene.

A new paper, “Long-Context State-Space Video World Models” by researchers from Stanford University, Princeton University, and Adobe Research, proposes an innovative solution to this challenge. They introduce a novel architecture that leverages State-Space Models (SSMs) to extend temporal memory without sacrificing computational efficiency.

The core problem lies in the quadratic computational complexity of attention mechanisms with respect to sequence length. As the video context grows, the resources required for attention layers explode, making long-term memory impractical for real-world applications. This means that after a certain number of frames, the model effectively “forgets” earlier events, hindering its performance on tasks that demand long-range coherence or reasoning over extended periods.

The authors’ key insight is to leverage the inherent strengths of State-Space Models (SSMs) for causal sequence modeling. Unlike previous attempts that retrofitted SSMs for non-causal vision tasks, this work fully exploits their advantages in processing sequences efficiently.

The proposed Long-Context State-Space Video World Model (LSSVWM) incorporates several crucial design choices:

1. Block-wise SSM Scanning Scheme: This is central to their design. Instead of processing the entire video sequence with a single SSM scan, they employ a block-wise scheme. This strategically trades off some spatial consistency (within a block) for significantly extended temporal memory. By breaking down the long sequence into manageable blocks, they can maintain a compressed “state” that carries information across blocks, effectively extending the model’s memory horizon.
2. Dense Local Attention: To compensate for the potential loss of spatial coherence introduced by the block-wise SSM scanning, the model incorporates dense local attention. This ensures that consecutive frames within and across blocks maintain strong relationships, preserving the fine-grained details and consistency necessary for realistic video generation. This dual approach of global (SSM) and local (attention) processing allows them to achieve both long-term memory and local fidelity.

The paper also introduces two key training strategies to further improve long-context performance:

• Diffusion Forcing: This technique encourages the model to generate frames conditioned on a prefix of the input, effectively forcing it to learn to maintain consistency over longer durations. By sometimes not sampling a prefix and keeping all tokens noised, the training becomes equivalent to diffusion forcing, which is highlighted as a special case of long-context training where the prefix length is zero. This pushes the model to generate coherent sequences even from minimal initial context.
• Frame Local Attention: For faster training and sampling, the authors implemented a “frame local attention” mechanism. This utilizes FlexAttention to achieve significant speedups compared to a fully causal mask. By grouping frames into chunks (e.g., chunks of 5 with a frame window size of 10), frames within a chunk maintain bidirectionality while also attending to frames in the previous chunk. This allows for an effective receptive field while optimizing computational load.

The researchers evaluated their LSSVWM on challenging datasets, including Memory Maze and Minecraft, which are specifically designed to test long-term memory capabilities through spatial retrieval and reasoning tasks.

The experiments demonstrate that their approach substantially surpasses baselines in preserving long-range memory. Qualitative results, as shown in supplementary figures (e.g., S1, S2, S3), illustrate that LSSVWM can generate more coherent and accurate sequences over extended periods compared to models relying solely on causal attention or even Mamba2 without frame local attention. For instance, on reasoning tasks for the maze dataset, their model maintains better consistency and accuracy over long horizons. Similarly, for retrieval tasks, LSSVWM shows improved ability to recall and utilize information from distant past frames. Crucially, these improvements are achieved while maintaining practical inference speeds, making the models suitable for interactive applications.

The Paper Long-Context State-Space Video World Models is on arXiv

The post Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models first appeared on Synced.

🔗 Sumber: syncedreview.com