MAROKO133 Breaking ai: The Mysterious Interstellar Object May Have Just Exploded Wajib Bac

📌 MAROKO133 Update ai: The Mysterious Interstellar Object May Have Just Exploded H

Mysterious interstellar object 3I/ATLAS recently made its closest pass of the Sun, or perihelion, brightening up in observations as solar radiation caused it to shed gases at an immense rate.

The object, widely believed to be a comet, is losing a staggering amount of mass as it reemerges from behind the Sun — so much, in fact, that Harvard astrophysicist and close 3I/ATLAS watcher Avi Loeb suggests that it may have just broken up into well over a dozen pieces.

New images taken by British astronomers Michael Buechner and Frank Niebling show the object growing a massive “anti-tail” and a separate, “smoking” trail, jets that extend approximately 620,000 miles towards the Sun and 1,860,000 miles in the opposite direction, respectively, as Loeb notes in a new blog post.

“For a natural comet, the outflow velocity of the jets is expected to be [0.248 miles] per second… at the distance of 3I/ATLAS from the Sun,” he added. “At that speed, the jets must have persisted over a timescale of 1–3 months.”

However, according to Loeb’s calculations, 3I/ATLAS would’ve needed to absorb an enormous amount of energy from the Sun in order to sublimate the copious amounts of carbon dioxide ice and water ice required to lose a such huge proportion of its mass.

“At its perihelion distance, the Sun provided 700 Joules per square meter per second,” Loeb wrote. “This means that the absorbing area of 3I/ATLAS must have been larger than [617 square miles],” roughly equivalent to a sphere with a diameter of 14.3 miles.

That’s four times as large as his previous estimate of the object measuring at least 3.1 miles across, with a mass of at least 33 billion tons.

While solar system comets are expected to shed mass as they approach the Sun, 3I/ATLAS still appears to be an outlier.

“The required surface area of 3I/ATLAS to provide the inferred mass loss from the latest post-perihelion image, is at least 16 times larger than the upper limit derived here from its Hubble image on July 21, 2025,” Loeb wrote. “When the Webb data was taken on August 6, 2025, 3I/ATLAS lost only [330 pounds] per second.”

In other words, the mysterious visitor went from shedding several hundred pounds a second in August to roughly 4.4 million pounds a second near its perihelion, a “dramatic increase,” per Loeb.

“Was the dramatic mass loss and brightening of 3I/ATLAS at perihelion evidence that it disintegrated?” the astronomer questioned. “Breakup into fragments would have increased the surface area of its material.”

Loeb suggests that 3I/ATLAS could have split up into “at least 16 equal pieces, and likely many more,” which “would mean that 3I/ATLAS exploded at perihelion and we are witnessing the resulting fireworks.”

Alternatively, the astronomer isn’t willing to rule out that 3I/ATLAS may be “something other than a natural comet” if “upcoming observations would reveal that 3I/ATLAS was not decimated by the Sun and maintained its integrity as a single body.”

Ever since its discovery, Loeb has proffered a series of eyebrow-raising suggestions about the object — most notably that it could be a giant alien spacecraft, citing its unusual properties ranging from its enormous suspected mass to its trajectory that’s closely aligned with the solar system’s planets.

“Technological thrusters require a much smaller mass loss in order to produce the observed jets around 3I/ATLAS,” he wrote in his latest blog. “Chemical rockets are propelled by an exhaust speed of [1.86 to 3.1 miles per second], which is ten times larger than the maximum ejection speed of volatiles sublimated by sunlight from natural cometary surfaces.”

“Alien-tech thrusters might employ yet higher exhaust speeds, reducing the required mass loss by several orders of magnitude and making the required fuel a small [proportion] of the spacecraft mass,” he suggested.

Fortunately, we‘ll have several more opportunities to get a closer look. 3I/ATLAS is expected to make its closest pass of the Earth on December 19 and Jupiter in March. How much of its nucleus will be left by that time remains to be seen.

More on 3I/ATLAS: NASA Withholding New Images of Mysterious Object From Beyond Solar System

The post The Mysterious Interstellar Object May Have Just Exploded appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Breaking ai: ByteDance Introduces Astra: A Dual-Model Architecture for

The increasing integration of robots across various sectors, from industrial manufacturing to daily life, highlights a growing need for advanced navigation systems. However, contemporary robot navigation systems face significant challenges in diverse and complex indoor environments, exposing the limitations of traditional approaches. Addressing the fundamental questions of “Where am I?”, “Where am I going?”, and “How do I get there?”, ByteDance has developed Astra, an innovative dual-model architecture designed to overcome these traditional navigation bottlenecks and enable general-purpose mobile robots.

Traditional navigation systems typically consist of multiple, smaller, and often rule-based modules to handle the core challenges of target localization, self-localization, and path planning. Target localization involves understanding natural language or image cues to pinpoint a destination on a map. Self-localization requires a robot to determine its precise position within a map, especially challenging in repetitive environments like warehouses where traditional methods often rely on artificial landmarks (e.g., QR codes). Path planning further divides into global planning for rough route generation and local planning for real-time obstacle avoidance and reaching intermediate waypoints.

While foundation models have shown promise in integrating smaller models to tackle broader tasks, the optimal number of models and their effective integration for comprehensive navigation remained an open question.

ByteDance’s Astra, detailed in their paper “Astra: Toward General-Purpose Mobile Robots via Hierarchical Multimodal Learning” (website: https://astra-mobility.github.io/), addresses these limitations. Following the System 1/System 2 paradigm, Astra features two primary sub-models: Astra-Global and Astra-Local. Astra-Global handles low-frequency tasks like target and self-localization, while Astra-Local manages high-frequency tasks such as local path planning and odometry estimation. This architecture promises to revolutionize how robots navigate complex indoor spaces.

Astra-Global: The Intelligent Brain for Global Localization

Astra-Global serves as the intelligent core of the Astra architecture, responsible for critical low-frequency tasks: self-localization and target localization. It functions as a Multimodal Large Language Model (MLLM), adept at processing both visual and linguistic inputs to achieve precise global positioning within a map. Its strength lies in utilizing a hybrid topological-semantic graph as contextual input, allowing the model to accurately locate positions based on query images or text prompts.

The construction of this robust localization system begins with offline mapping. The research team developed an offline method to build a hybrid topological-semantic graph G=(V,E,L):

• V (Nodes): Keyframes, obtained by temporal downsampling of input video and SfM-estimated 6-Degrees-of-Freedom (DoF) camera poses, act as nodes encoding camera poses and landmark references.
• E (Edges): Undirected edges establish connectivity based on relative node poses, crucial for global path planning.
• L (Landmarks): Semantic landmark information is extracted by Astra-Global from visual data at each node, enriching the map’s semantic understanding. These landmarks store semantic attributes and are connected to multiple nodes via co-visibility relationships.

In practical localization, Astra-Global’s self-localization and target localization capabilities leverage a coarse-to-fine two-stage process for visual-language localization. The coarse stage analyzes input images and localization prompts, detects landmarks, establishes correspondence with a pre-built landmark map, and filters candidates based on visual consistency. The fine stage then uses the query image and coarse output to sample reference map nodes from the offline map, comparing their visual and positional information to directly output the predicted pose.

For language-based target localization, the model interprets natural language instructions, identifies relevant landmarks using their functional descriptions within the map, and then leverages landmark-to-node association mechanisms to locate relevant nodes, retrieving target images and 6-DoF poses.

To empower Astra-Global with robust localization abilities, the team employed a meticulous training methodology. Using Qwen2.5-VL as the backbone, they combined Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO). SFT involved diverse datasets for various tasks, including coarse and fine localization, co-visibility detection, and motion trend estimation. In the GRPO phase, a rule-based reward function (including format, landmark extraction, map matching, and extra landmark rewards) was used to train for visual-language localization. Experiments showed GRPO significantly improved Astra-Global’s zero-shot generalization, achieving 99.9% localization accuracy in unseen home environments, surpassing SFT-only methods.

Astra-Local: The Intelligent Assistant for Local Planning

Astra-Local acts as the intelligent assistant for Astra’s high-frequency tasks, a multi-task network capable of efficiently generating local paths and accurately estimating odometry from sensor data. Its architecture comprises three core components: a 4D spatio-temporal encoder, a planning head, and an odometry head.

The 4D spatio-temporal encoder replaces traditional mobile stack perception and prediction modules. It begins with a 3D spatial encoder that processes N omnidirectional images through a Vision Transformer (ViT) and Lift-Splat-Shoot to convert 2D image features into 3D voxel features. This 3D encoder is trained using self-supervised learning via 3D volumetric differentiable neural rendering. The 4D spatio-temporal encoder then builds upon the 3D encoder, taking past voxel features and future timestamps as input to predict future voxel features through ResNet and DiT modules, providing current and future environmental representations for planning and odometry.

The planning head, based on pre-trained 4D features, robot speed, and task information, generates executable trajectories using Transformer-based flow matching. To prevent collisions, the planning head incorporates a masked ESDF loss (Euclidean Signed Distance Field). This loss calculates the ESDF of a 3D occupancy map and applies a 2D ground truth trajectory mask, significantly reducing collision rates. Experiments demonstrate its superior performance in collision rate and overall score on out-of-distribution (OOD) datasets compared to other methods.

The odometry head predicts the robot’s relative pose using current and past 4D features and additional sensor data (e.g., IMU, wheel data). It trains a Transformer model to fuse information from different sensors. Each sensor modality is processed by a specific tokenizer, combined with modality embeddings and temporal positional embeddi…

Konten dipersingkat otomatis.

🔗 Sumber: syncedreview.com