MAROKO133 Eksklusif ai: New sodium-sulfur battery design from China pushes energy density

📌 MAROKO133 Update ai: New sodium-sulfur battery design from China pushes energy d

Researchers at Shanghai Jiao Tong University in China have designed a new sodium-sulfur battery with higher power density and discharge capacity than before, enabling a cheaper, safer alternative to lithium-ion batteries. 

Our switch to electric energy has driven increased demand for energy storage devices. Lithium-ion batteries are the most energy-dense solutions we currently know and are widely used.

However, issues of thermal runaway and fire risks prevent their use in large-scale applications. Moreover, the increased demand has driven up lithium prices, making energy storage more expensive by the day. 

Researchers are looking for alternatives to replace lithium-ion batteries. Sodium is highly abundant and can serve as a low-cost alternative, and is part of multiple research projects globally where sodium is used in various combinations. 

Researchers at Shanghai Jiao Tong University teamed up sodium with sulfur to make a high-energy-density battery. 

Problems with Sodium-sulfur batteries

This is not the first attempt to pair sodium and sulfur. Batteries made using Na–S or S/Na2S chemistry required large amounts of sodium but delivered low voltage. The redox reaction with sulfur generates 4 valence electrons, producing a voltage of 3.6V, but replicating this reaction at room temperature was a major challenge. 

According to the researchers, the S/Na2S conversion reaction at the cathode yields a limited discharge voltage of less than 1.6 V, which is far lower than that of Li-ion counterparts.

To overcome this, large amounts of sodium must be used at the anode, which can be 10 times or more than in a lithium battery. This defeats the purpose of using a cheaper material and also affects energy and power densities. 

Switching to S0/S4+ redox chemistry 

The researchers unlocked the sodium-sulfur battery puzzle by switching to S0/S4+ redox chemistry and created high-voltage anode-free batteries. This design consists of an aluminum (Al) foil anode current collector, an S8 cathode, sodium dicyanamide (NaDCA) in a non-flammable chloroaluminate electrolyte, separated with a glass fiber.

According to the researchers, the dicyanamide anion in the electrolyte helps unlock S/SCl4 chemistry at the cathode while also improving the reversibility of sodium plating/stripping at the anode.

The improved performance of this design is evident in its maximum energy density of 1,198 Wh/kg, discharge capacity of 715 mAh g−1, and power density of 23,773 W/kg. 

When the researchers added a Bi-COF catalyst at the cathode, the discharge capacity further increased to 1,206 mAh/g, while the energy density rose to 2,021 Wh/kg.

With an estimated cost of $5.03 per kWh, the sodium-sulfur battery costs an order of magnitude less than its lithium counterparts. Safety is inherently enhanced because the electrolyte is non-flammable. 

However, researchers still need to work on a few issues before the battery design can be commercially available. The electrolyte, though non-flammable, is highly corrosive and difficult to handle. Additionally, it is only stable in the short term when exposed to air, while long-term stability is currently unknown. 

The team is, however, confident that these issues can be addressed and that they will help improve device safety across devices ranging from wearables to grid-scale energy storage. 

The research findings were published in the journal Nature. 

🔗 Sumber: interestingengineering.com

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

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