Search AI Products and News

Google recently released a new version of its Gemma3 series, exciting many AI enthusiasts. Just a month after its initial launch, Google released a Quantization Aware Training (QAT) optimized version of Gemma3, aiming to significantly reduce memory requirements while maintaining model quality. Specifically, the QAT-optimized Gemma3 27B model reduces VRAM requirements from 54GB to 14.1GB, meaning users can now run it on a single NVIDIA RTX 3090.

Google recently released a groundbreaking AI model, DolphinGemma, aimed at understanding the communication patterns of wild Atlantic spotted dolphins. Developed in collaboration with the Wild Dolphin Project (WDP) and researchers from the Georgia Institute of Technology, the project leverages WDP's extensive database of dolphin audio and video recordings accumulated over nearly 40 years in the Bahamas. Trained on this valuable data, DolphinGemma utilizes Google's advanced SoundStream tokenizer to segment...

Google recently announced the launch of its new open-source model, TxGemma, designed to significantly improve the efficiency of therapeutic drug development. Developed by the Google DeepMind team, this model is a fine-tuned version of their advanced Gemma model family, incorporating powerful natural language understanding, scientific prediction, and multi-turn dialogue capabilities. TxGemma aims to revolutionize drug discovery. Its release is considered a significant advancement in the application of AI to biomedicine, promising to drastically shorten the time it takes for drugs to move from the lab to clinical trials, while also reducing costs.

Researchers at the Hebrew University of Jerusalem recently discovered that in Retrieval Augmented Generation (RAG) systems, the number of documents processed significantly impacts language model performance, even when the total text length remains constant. The research team conducted experiments using 2,417 questions from the MuSiQue validation dataset, each linked to 20 Wikipedia paragraphs. Two to four paragraphs contained relevant answer information, with the remaining paragraphs serving as distractors. To study the impact of the number of documents, the team created multiple data partitions, progressively reducing the number of documents from 20 to...

Reka AI, founded by a dozen former Google DeepMind scientists, has unveiled its first open-source model: Reka Flash 3. This 21-billion parameter inference model has garnered significant attention. Despite its relatively smaller parameter count, Reka Flash 3 is a general-purpose reasoning model trained from scratch. It underwent supervised fine-tuning on synthetic and public datasets and further refinement through model-based techniques.

Google CEO Sundar Pichai announced at a launch event that Google has open-sourced its latest multimodal large model, Gemma-3. This model is attracting significant attention for its low cost and high performance. Gemma-3 offers four different parameter scale options: 1 billion, 4 billion, 12 billion, and 27 billion parameters. Surprisingly, the largest 27 billion parameter model only requires a single H100 GPU for efficient inference, while similar models often require ten times the computing power.

Recently, Google announced the release of a brand new Vision-Language Model (VLM) called PaliGemma 2 Mix. This model combines image processing and natural language processing capabilities, allowing it to understand visual information and text input simultaneously, generating corresponding outputs as needed. This marks a significant breakthrough in artificial intelligence technology for multi-task processing. PaliGemma 2 Mix boasts powerful features, integrating image description and optical character recognition.

In recent years, the rapid development of artificial intelligence (AI) technology, especially driven by countries like the United States and China, has made AI a focal point of global technological competition. To this end, the EU recently announced a funding program of up to 200 billion euros to promote development in this field. Despite Germany's strong capabilities in AI research, why hasn't a single successful AI enterprise emerged? Source: Image generated by AI, image courtesy of service provider Midjourney according to Professor Bjorn from the University of Munich.

Since last summer, billionaire Elon Musk's artificial intelligence company xAI has promised that its flagship model Grok3 would be released by the end of 2024, in response to models like OpenAI's GPT-4 and Google's Gemini. However, with January 2nd now passed, Grok3 has yet to appear, nor are there any signs of an upcoming release. Musk stated on the X social platform in July that Grok3 would be launched by the end of the year after training on 100,000 H100 GPUs, representing a 'significant leap'.

An unexpected personnel change brought OpenAI's Chief Product Officer Kevin Weil to the stage of the Marriott Hotel in San Francisco. This Silicon Valley product master, who previously achieved commercial miracles at Twitter and Instagram, now bears the responsibility of OpenAI's commercialization transformation. Just days earlier, the sudden departure of the company's CTO Mira Murati once again placed OpenAI in the spotlight. This executive reshuffle is not a coincidence. Over the past year, OpenAI has faced wave after wave of challenges.

IBM has officially launched its next-generation open-source large language model, Granite3.1, aiming to secure a leading position in the enterprise AI sector. This series of models features an extended context length of 128K, embedding models, built-in hallucination detection capabilities, and significant performance improvements. IBM claims that the Granite8B Instruct model outperforms its open-source competitors of similar scale, including Meta's Llama3.1, Qwen2.5, and Google's Gemma2.