MAROKO133 Breaking ai: 2,900-year-old Siberian site produces earliest ‘Scythian animal sty

📌 MAROKO133 Update ai: 2,900-year-old Siberian site produces earliest ‘Scythian an

A new study sheds new light on one of the ancient world’s most iconic yet elusive artistic symbols: the “Scythian animal style.”

Study authors reported in Antiquity that they discovered the earliest examples of the “Scythian animal style” at Tunnug 1, a royal tomb in Siberia, Russia. Its remarkable age provides unprecedented evidence regarding ancient Scythian culture. The tomb dates back to the 9th century BCE.

“This is really a once-in-a-lifetime opportunity,” lead researcher Dr. Gino Caspari told Phys, describing Tunnug-1 as “extremely rare, if not unique.” The royal tomb yielded the oldest-known Scythian animal motifs.

The Scythian animal style is characterized by a distinctive and dynamic artistic tradition featuring stylized animal depictions rendered in gold, bronze, and other materials.

The study authors noted that while “the literature on Scythian animal style is extensive,” reflecting over a century of scholarly effort to trace its origins, archaeologists lacked a site like Tunnug-1 to securely confirm its age.

“Tunnug 1 stands as one of the earliest monuments where the full expression of the “Scythian triad”—weapons, horse tack, animal-style motifs—is attested,” the study authors continued.

Though hypotheses for the Scythian animal style’s indigenous origins have long existed, the collection from Tunnug offers the first securely dated contextual evidence of its emergence in “early nomadic cultures.”

These artifacts “offer a rare glimpse into the genesis of a visual system that would later spread across the steppe and become emblematic of Scythian identity.”

The beginning of a distinctly eastern artistic tradition

The importance of Tunnug 1 extends beyond the extraordinary collection of artifacts.

Located in Siberia’s “Valley of the Kings,” this large royal tomb has shifted the understanding of how Scythian culture developed across the steppe—the vast grassland ecoregion stretching from Hungary to China, known for its role in the history of nomadic cultures.

Tunnug 1 confirms the hypothesis that the culture originated in Siberia, not western Europe. This confirms that the Scythian animal style first developed in the east, later expanding its iconography and symbolic significance to include deer, griffins, boars, and more. Even these early examples, nearly 3,000 years old, had royal ties, demonstrating their immediate representation of status, cosmology, and identity.

However, the collection at Tunnug 1 reflects a simpler beginning. Archaeologists uncovered only four animal motifs—rams, felines, birds, and snakes—suggesting the artistic tradition started on a more modest note.

Dr. Caspari added, “The steppe has to be considered prehistoric,” meaning the symbolism cannot be definitively interpreted. However, “clearly,” he said, “wild animals are very important as spirits inhabiting the natural world.” Notably, the depictions are of wild animals, not domesticated ones, even though the region began herding animals quite early, around 4000 BCE, as per Britannica.

A style that many riffed off

The Scythian animal style would eventually flourish, becoming an umbrella term encompassing an extensive range of animals, including horses (which were notably domesticated in that region). The smaller, bronze set found at Tunnug 1 indicates a modest debut, particularly since the style would later become strongly associated with gold.

Having analyzed the treasure, the study authors now postulate that the diversity that would come to define the Scythian animal style might reflect different social groups utilizing and adapting the theme for funerary rituals.

“At the moment, we are mostly working with data we’ve already retrieved because analyzing these and properly publishing the results takes time. People can definitely look forward to additional papers coming out soon,” concluded Dr. Caspari.

The study has been published in the journal Antiquity.

🔗 Sumber: interestingengineering.com

📌 MAROKO133 Hot ai: Adobe Research Unlocking Long-Term Memory in Video World Model

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