MAROKO133 Eksklusif ai: Subsurface map of Antarctica reveals hidden terrain with the secre

📌 MAROKO133 Hot ai: Subsurface map of Antarctica reveals hidden terrain with the s

What lies beneath Antarctica’s immense and impenetrable ice sheet? Until now, we didn’t fully know, but a new, groundbreaking study reached deeper than any other to discover a hidden world that might hold the key to predicting the future of the continent, as well as our world.

In a new, groundbreaking study published in Science, authors explained that “less is known about the topography beneath the ice of Antarctica than any other planetary surface in the inner solar system.”

As the least studied region in the known universe, this mysterious area, cloaked by the Antarctic Ice Sheet, “offers critical insights into its geological history and influences how the ice reacts to climate changes,” as study authors continued.

Scientists ventured into our planet’s icy frontier with satellite data and physics. They deciphered the complex movements of the ice as if reading a secret language, creating a map of a subglacial wonderland of mountains, deep canyons, and rugged hills, as per the BBC.

And this sub-surface map of Antarctica would aid climate scientists in understanding how glaciers will move as the ice continues to melt under the threat of climate change. The urgency surrounding the speed of their disappearance is one of the most significant uncertainties confronting us today.

Did these scientists just break new ground?

Is Narnia actually under Antarctica?

Antarctica has long sparked the imagination, characterized as the Earth’s southernmost, coldest, and driest polar desert. Ancient glaciers, expansive ice shelves, and towering mountain ranges create an awe-inspiring landscape. It also serves as a sanctuary for unique wildlife, including penguins and seals, while being a hub for extensive international scientific research.

Although previous surveys have provided insights, a lead climate scientist said to the BBC that the latest map was “a really useful product” because it fills crucial gaps that have persisted for decades. No researcher had penetrated these depths or mapped the ice sheet’s underlying terrain in such vivid, thrilling detail.

The research team analyzed the surface using high-resolution satellite imagery, followed by a technique in physics known as Ice Flow Perturbation Analysis (IFPA). They traced the landscape by studying how the ice moved around it.

Beneath a colossal 5.4 million square miles of ice sheet, the study documented a staggering 71,997 hills and mapped a valley that stretched 248.5 miles within the Maud Subglacial Basin. Beneath Antarctica, river channels stretched hundreds of miles, and some regions even evoked alpine landscapes, Live Science reported. Detected transitions between highland plateaus and low-lying basins revealed tectonic boundaries that hint at the dynamic geological processes at play.

Indeed, an entire world awaited researchers beneath the surface that exercised a direct influence on the surface.

The key to climate change

The wealth of valuable information unearthed by the study stands to become as immense in significance as the map, as the rapidly melting Antarctic ice remains one of the era’s biggest concerns.

As the BBC elaborated, this hidden topography will significantly influence the movement of glaciers. But most importantly, the speed at which they might disappear and displace themselves.

With the map, scientists can improve their predictions on how swiftly the Antarctic ice will contribute to global sea-level rise.

Read the study in Science.

🔗 Sumber: interestingengineering.com

📌 MAROKO133 Update ai: MIT Researchers Unveil “SEAL”: A New Step Towards Self-Impr

The concept of AI self-improvement has been a hot topic in recent research circles, with a flurry of papers emerging and prominent figures like OpenAI CEO Sam Altman weighing in on the future of self-evolving intelligent systems. Now, a new paper from MIT, titled “Self-Adapting Language Models,” introduces SEAL (Self-Adapting LLMs), a novel framework that allows large language models (LLMs) to update their own weights. This development is seen as another significant step towards the realization of truly self-evolving AI.

The research paper, published yesterday, has already ignited considerable discussion, including on Hacker News. SEAL proposes a method where an LLM can generate its own training data through “self-editing” and subsequently update its weights based on new inputs. Crucially, this self-editing process is learned via reinforcement learning, with the reward mechanism tied to the updated model’s downstream performance.

The timing of this paper is particularly notable given the recent surge in interest surrounding AI self-evolution. Earlier this month, several other research efforts garnered attention, including Sakana AI and the University of British Columbia’s “Darwin-Gödel Machine (DGM),” CMU’s “Self-Rewarding Training (SRT),” Shanghai Jiao Tong University’s “MM-UPT” framework for continuous self-improvement in multimodal large models, and the “UI-Genie” self-improvement framework from The Chinese University of Hong Kong in collaboration with vivo.

Adding to the buzz, OpenAI CEO Sam Altman recently shared his vision of a future with self-improving AI and robots in his blog post, “The Gentle Singularity.” He posited that while the initial millions of humanoid robots would need traditional manufacturing, they would then be able to “operate the entire supply chain to build more robots, which can in turn build more chip fabrication facilities, data centers, and so on.” This was quickly followed by a tweet from @VraserX, claiming an OpenAI insider revealed the company was already running recursively self-improving AI internally, a claim that sparked widespread debate about its veracity.

Regardless of the specifics of internal OpenAI developments, the MIT paper on SEAL provides concrete evidence of AI’s progression towards self-evolution.

Understanding SEAL: Self-Adapting Language Models

The core idea behind SEAL is to enable language models to improve themselves when encountering new data by generating their own synthetic data and optimizing their parameters through self-editing. The model’s training objective is to directly generate these self-edits (SEs) using data provided within the model’s context.

The generation of these self-edits is learned through reinforcement learning. The model is rewarded when the generated self-edits, once applied, lead to improved performance on the target task. Therefore, SEAL can be conceptualized as an algorithm with two nested loops: an outer reinforcement learning (RL) loop that optimizes the generation of self-edits, and an inner update loop that uses the generated self-edits to update the model via gradient descent.

This method can be viewed as an instance of meta-learning, where the focus is on how to generate effective self-edits in a meta-learning fashion.

A General Framework

SEAL operates on a single task instance (C,τ), where C is context information relevant to the task, and τ defines the downstream evaluation for assessing the model’s adaptation. For example, in a knowledge integration task, C might be a passage to be integrated into the model’s internal knowledge, and τ a set of questions about that passage.

Given C, the model generates a self-edit SE, which then updates its parameters through supervised fine-tuning: θ′←SFT(θ,SE). Reinforcement learning is used to optimize this self-edit generation: the model performs an action (generates SE), receives a reward r based on LMθ′’s performance on τ, and updates its policy to maximize the expected reward.

The researchers found that traditional online policy methods like GRPO and PPO led to unstable training. They ultimately opted for ReST^EM, a simpler, filtering-based behavioral cloning approach from a DeepMind paper. This method can be viewed as an Expectation-Maximization (EM) process, where the E-step samples candidate outputs from the current model policy, and the M-step reinforces only those samples that yield a positive reward through supervised fine-tuning.

The paper also notes that while the current implementation uses a single model to generate and learn from self-edits, these roles could be separated in a “teacher-student” setup.

Instantiating SEAL in Specific Domains

The MIT team instantiated SEAL in two specific domains: knowledge integration and few-shot learning.

• Knowledge Integration: The goal here is to effectively integrate information from articles into the model’s weights.
• Few-Shot Learning: This involves the model adapting to new tasks with very few examples.

Experimental Results

The experimental results for both few-shot learning and knowledge integration demonstrate the effectiveness of the SEAL framework.

In few-shot learning, using a Llama-3.2-1B-Instruct model, SEAL significantly improved adaptation success rates, achieving 72.5% compared to 20% for models using basic self-edits without RL training, and 0% without adaptation. While still below “Oracle TTT” (an idealized baseline), this indicates substantial progress.

For knowledge integration, using a larger Qwen2.5-7B model to integrate new facts from SQuAD articles, SEAL consistently outperformed baseline methods. Training with synthetically generated data from the base Qwen-2.5-7B model already showed notable improvements, and subsequent reinforcement learning further boosted performance. The accuracy also showed rapid improvement over external RL iterations, often surpassing setups using GPT-4.1 generated data within just two iterations.

Qualitative examples from the paper illustrate how reinforcement learning leads to the generation of more detailed self-edits, resulting in improved performance.

While promising, the researchers also acknowledge some limitations of the SEAL framework, including aspects related to catastrophic forgetting, computational overhead, and context-dependent evaluation. These are discussed in detail in the original paper.

Original Paper: https://arxiv.org/pdf/2506.10943

Project Site: https://jyopari.github.io/posts/seal

Github Repo: https://github.com/Continual-Intelligence/SEAL

The post MIT Researchers Unveil “SEAL”: A New Step Towards Self-Improving AI first appeared on Synced.

🔗 Sumber: syncedreview.com