📌 TOPINDIATOURS Hot ai: China’s method delivers 3 times more olefin production fro

Chinese researchers have developed a new method for creating coal-sourced chemicals that reduces carbon dioxide emissions while increasing yield and efficiency. In fact, they found that their new process can result in three times as much useful product.

This, the researchers claim, is achieved using a relatively simple modification to the underlying chemical process. China remains heavily reliant on coal-to-chemical processes, particularly for producing olefins.

In case you are unaware, these are compounds made of hydrogen and carbon that typically contain one or more pairs of carbon atoms linked by a double bond. Also called alkenes, the most important examples in industry include ethylene and propylene.

To make them, coal feed is usually converted to something called syngas (carbon monoxide and hydrogen gas). This, in turn, is then converted to methanol and then finally to olefins.

More olefins, fewer carbon emissions

In other parts of the world, steam is used to convert oil or natural gas to make the same product, which has fewer steps but is equally inefficient in terms of energy and waste.

China primarily uses coal as it has an abundant domestic supply and tends to prefer to be less reliant on importing oil and gas. Olefins are high-value chemicals used to make things like plastics (polyethylene, polypropylene), pharmaceuticals, and other advanced materials.

Normally, converting coal to chemicals produces a lot of waste carbon dioxide and has other side effects that lower efficiency.

In processes like this, any loss of carbon as carbon dioxide is a key sign that the underlying process is inefficient. Any lost carbon as a gas like this ultimately means that it is no longer part of the final product, whatever that may be.

These inefficiencies tend to result in less useful product per ton of feed coal. To address this, the researchers performed what they have termed a “molecular switch.”

In practical terms, this means that they altered the reaction pathway. They were also able to block the chemical side reactions that normally waste carbon, create carbon dioxide, and reduce yield.

Boosting output and protecting the planet

To this end, valuable carbon isn’t lost as an exhaust gas but rather is retained within the product molecules and final chemicals. By keeping more carbon in the final product, the new process not only cuts emissions but also increases yield and ultimately improves its economics.

It is important to note that this new process doesn’t suddenly make fossil fuels, like coal, now “green.” The process is more of a coal-to-product process efficiency boost that means more of the raw material is actually converted to the final product.

However, it does show that concerns around reducing pollution don’t have to come at the cost of industrial efficiency. Looking beyond the production of olefins from coal, it is yet to be seen if such a process could be used for energy production from fossil fuels.

It could, for example, help improve the process of converting coal, oil, or gas into cleaner intermediate fuels for burning. Ultimately, however, the inherent inefficiencies of combustion will likely always remain an issue.

🔗 Sumber: interestingengineering.com

📌 TOPINDIATOURS Breaking ai: MIT Researchers Unveil “SEAL”: A New Step Towards Sel

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