Dow Jones, the parent company of The Wall Street Journal, and the New York Post are suing AI-driven search startup Perplexity for using their news content to train its large language model. The publications from these two news groups accuse Perplexity of copyright infringement by generating answers to people's queries using their articles, thereby diverting traffic from the publications' websites. 'This lawsuit is brought by news publishers seeking compensation as Perplexity openly plans to provide services for free.'

A recent study led by the Technion-Israel Institute of Technology reveals that large language models (LLMs) may be more knowledgeable than they appear. Researchers found that LLMs' internal representations encode information about the correctness of their outputs, allowing them to identify the correct answer even when they ultimately generate an incorrect one. The research team focused on analyzing LLM errors in long text generation, which is more aligned with their real-world application scenarios. They constructed an error detection dataset to compare the model's outputs.

In the episode of HBR Live: The Future of Business, Baidu founder Robin Li had an in-depth conversation with Adi Ignatius, editor-in-chief of the Harvard Business Review, discussing the development trends of artificial intelligence (AI), the phenomenon of bubbles, and the future relationship between AI and humanity. Robin Li believes that although it is difficult to avoid bubbles in the early stages of technological waves, ultimately, 1% of companies will emerge after the bubble and create tremendous value for society.

Recently, teams from Cambridge University and others published a groundbreaking paper that reveals the true nature of large models (LLMs), providing an in-depth analysis of the actual performance of current large language models (LLMs). The results are shocking — these AI models, once highly anticipated, perform far below expectations on many basic tasks. The study conducted a comprehensive evaluation of several cutting-edge models, including o1-preview. The findings indicate significant discrepancies in understanding capabilities between AI models and humans. Surprisingly, the models struggle with tasks that humans excel at.

Tencent Youtu Lab has collaborated with the research team at Shanghai Jiao Tong University to introduce a revolutionary knowledge enhancement method, paving new avenues for optimizing large models. This innovative technology discards the limitations of traditional model fine-tuning, directly extracting knowledge from open-source data, significantly simplifying the model optimization process, and achieving outstanding performance that surpasses existing state-of-the-art technologies (SOTA) across multiple tasks. In recent years, while large language models (LLMs) have made significant progress in various fields, they still face numerous challenges in practical applications. Traditional model fine-tuning methods...

At the 2nd Cybersecurity (Tianjin) Forum held on September 4, 2024, Zhou Hongyi, the founder of 360 Group, delivered a speech titled 'Building a Security Large Model, Reshaping Security with AI', sharing his insights on the cybersecurity landscape in the era of large AI models and introducing 360 Group's practices and achievements in this field.

The new feature 'Model Arena' of 360AI Assistant allows users to choose from multiple large models, input the same prompt, and intuitively compare the outputs of different models. By evaluating model performance through a rich question bank, it enhances objectivity and comprehensiveness. This feature is combined with the new AI Assistant floating button in the 360 Browser, integrating products from 15 top large model companies, providing a platform for free selection and comparison, thereby enriching the user experience.

Language models have shown widespread applications across various fields, from education to legal consulting, and even in predicting medical risks. However, as these models gain more weight in decision-making processes, they may unintentionally reflect the biases present in the human training data, exacerbating discrimination against minority groups. Research has found that language models exhibit implicit racism, particularly in their treatment of African American English (AAE), demonstrating harmful dialect discrimination that is more negative than any known stereotypes against African Americans. The 'masking disguise' method was used to compare AAE with Standard American English.

The upcoming iPhone 17 series from Apple is expected to feature 12GB RAM to meet the memory demands of future AI large models. The current iPhone 15 Pro and iPhone 16 series with 8GB RAM are insufficient for supporting large models with around 7 billion parameters. A model with 13 billion parameters requires at least 5GB of memory, and after the operating system and applications take up space, 8GB RAM fails to provide enough room. The 12GB RAM configuration could offer ample space for larger AI models.

According to the evaluation results from the China Academy of Information and Communications Technology, Baidu Smart Cloud's Keway intelligent customer service product received the highest rating in the first batch of 'Intelligent Customer Service Based on Large Models,' marking authoritative recognition of Baidu's technical strength and application practices in this field. Intelligent customer service, as an important application of artificial intelligence, holds significant importance for industries such as finance, government, and telecommunications, helping institutions achieve cost reduction and efficiency enhancement. Keway, as the country's first intelligent customer service product based on large models, offers advantages such as comprehensive functionality, superior performance, and mature systems, efficiently solving complex problems and enhancing problem-solving autonomy.

The release of PaddlePaddle Framework 3.0 features a core upgrade to unified automatic parallelism technology, aimed at simplifying the distributed training process for large models and improving development efficiency. The new version supports mixed parallelism techniques from four-dimensional to five-dimensional, employing various parallel methods such as data parallelism, tensor model parallelism, pipeline parallelism, and grouped parameter slicing parallelism, effectively enhancing the training efficiency of large models. The automatic parallelism technology uses tensor slicing syntax to automatically infer distributed slicing states and add communication operators, reducing the difficulty of development. The principles of automatic parallelism include distributed tensor representation and slicing inference.

Tencent Youtu Lab and Shanghai Jiao Tong University released a comprehensive review focusing on dataset evaluation and selection for instruction tuning to enhance large model performance. The study aims to help large models grasp the essence of natural language processing by analyzing data quality, diversity, and importance to guide the construction of efficient and critical datasets. It highlights that data quality directly impacts tuning effect....

Mercedes-Benz (China) Investment Co., Ltd. has recently signed a strategic cooperation memorandum with Beijing Volcano Engine Technology Co., Ltd., announcing in-depth collaboration and exploration in the fields of large models, generative artificial intelligence, and big data technology. This strategic partnership marks a further acceleration of Mercedes-Benz's digital innovation in China, aiming to enhance the smart cockpit experience through AI technology and provide Chinese customers with a more convenient and personalized digital luxury experience.

The Tencent Hunyuan Large Model excelled in the August rankings of the SuperCLUE-V evaluation benchmark for Chinese multimodal large models, securing the top position among domestic large models and placing in the 'Exemplary Leaders' quadrant. Multimodal understanding requires the model to accurately identify image elements, comprehend their relationships, and generate natural language descriptions, testing the model's precision in image recognition and its understanding of the complex real world.

Long context understanding is a key challenge in the field of natural language processing, especially when large language models (LLMs) handle text that exceeds their context window size. To address this issue, researchers developed the LooGLE benchmark test, aimed at assessing LLMs' capability to understand extremely long documents (averaging 19.3k words, comprising 776 articles across multiple domains). LooGLE includes 7 tasks that cover both short and long dependencies, evaluating model understanding across texts of varying lengths. The test data is sourced from open-access content after 2022.

Zhang Xiangyu, an outstanding scientist in the field of AI, rose to fame for his involvement in the ResNet research, which innovatively introduced residual connections and is crucial for modern deep learning models. As a member of the post-90s generation, he joined Jietuo Star, sparking industry attention. Zhang's academic achievements include achieving the best historical result for Xi'an Jiaotong University in a mathematical modeling competition and co-completing the ResNet research during his time at Microsoft Research Asia. The ResNet paper has been cited over 230,000 times, making it one of the most popular deep learning works of this century.