# Nobel Prize in Physics Awarded for Machine Learning and Neural Networks

- The **Nobel Prize in Physics 2024** recognizes **John J. Hopfield** and **Geoffrey E. Hinton** for their pivotal contributions to **machine learning** through **artificial neural networks**, which have transformed computational capabilities.

- **Diff Transformer** enhances attention by focusing on relevant context while effectively canceling noise, utilizing a differential attention mechanism that computes scores as the difference between two softmax maps.

- **LongWriter** is a groundbreaking model capable of generating **over 10,000 words in just one minute**, utilizing advanced long-context LLMs, and is now deployable via [vllm](https://github.com/vllm-project/vllm) for enhanced performance.

- **Magic Kernel Sharp** is a superior image resizing algorithm, utilized by major platforms like **Facebook** and **Instagram**, enhancing image quality while improving CPU and storage efficiency since its inception in 2013.

- **Model2Vec** distills Sentence Transformer models into **30mb static embeddings** that are **up to 500x faster** than their original counterparts, enabling efficient CPU usage without requiring extensive hardware.

- **Canon's new FPA-1200NZ2C nanoimprint lithography machine aims to revolutionize chipmaking** by offering a low-cost alternative to ASML's expensive EUV technology, potentially reducing production costs by **one digit** and energy consumption by **up to 90%**.

- **Geoffrey Hinton** and **John Hopfield** received the **2024 Nobel Prize in Physics** for their groundbreaking work on **artificial neural networks**, which mimic human memory storage and retrieval, significantly advancing AI technology.

- The app developed for **GPT-2** extracts interpretable **"circuits"**, revealing how information flows through the model and demonstrating the formation of features that identify grammatical patterns.

- The **L-Mul algorithm** approximates floating-point multiplication using integer addition, achieving **higher precision** with significantly lower computational costs, potentially reducing energy consumption by **95%** for element-wise operations and **80%** for dot products.

- The **SWE-bench Multimodal** dataset introduces **617 new task instances** where AI agents tackle real bugs from **17 JavaScript GitHub repositories**, each accompanied by an image, enhancing the challenge for existing models.

- **Automated data cleaning** can significantly enhance model performance, as demonstrated by pruning the Pile dataset to 0.72% of its original size while achieving only a minor drop in benchmark scores for BERT and T5 models.

- **NVIDIA's cuLitho platform** is revolutionizing semiconductor manufacturing at TSMC by enabling accelerated computational lithography, which significantly enhances chip production efficiency and performance.

- **TableRAG** introduces a **Retrieval-Augmented Generation (RAG)** framework that enhances language models' ability to understand tables by efficiently pinpointing crucial information through **query expansion** and **schema retrieval**.

- **Foxconn's new supercomputer**, utilizing NVIDIA’s **GB200 NVL72 platform**, aims to achieve over **90 exaflops** of AI performance, positioning it as Taiwan's fastest and a significant player in global AI advancements.