LLM Pulse - July 01, 2024

New Releases 

1. Meta’s LLM Compiler is the latest AI breakthrough to change the way we code:

Meta has unveiled the Meta Large Language Model (LLM) Compiler, a suite of robust, open-source models designed to optimize code and revolutionize compiler design. Read More 

2. Scientist.com Unveils Elisa: An LLM -Powered Research Assistant for Scientists in the Pharmaceutical and Biotechnology Industries:

Scientist.com, the world's leading marketplace for biopharma research, today announced the launch of a new large language model (LLM)-powered research assistant called Elisa. Read More 

3. Salesforce Announces the World's First LLM Benchmark for CRM:

New benchmark and leaderboard give businesses the guidance they need to make smart decisions when evaluating generative AI models for their CRM systems. Read More 

4. Meta AI chatbot, powered by Meta Llama 3 LLM , comes to India across WhatsApp, Facebook, Messenger, Instagram, and Meta ai:

Meta announced that its Meta AI chatbot, powered by the company's most advanced Llama 3 large language model (LLM), is coming to India in English, as users can access it through WhatsApp, Facebook, Messenger, Instagram, and meta.ai. Read More 

5. Core42's Latest Arabic Large Language Model Now Available on Microsoft Azure:

Core42, a G42 company and full-spectrum AI enablement solutions provider, is excited to announce the availability of its latest Arabic Large Language Model (LLM), JAIS 30B Chat, on Microsoft Azure. Read More 

6. Google creates Tx-LLM for drug discovery and therapeutic development:

Google Research and Google DeepMind recently released a paper announcing the creation of a new LLM for drug discovery and therapeutic development dubbed Tx-LLM, fine-tuned from PaLM-2. Read More 

7. Oracle Corporation - Early LLM serving experience and performance results with AMD: 

The AMD Instinct MI300X GPUs are new entrant to the generative AI market and are proving to be strong contender for customers looking to run larger sized LLM model inference and training workloads. Read More 

8. Akkio Announces Launch of Agency Data LLM 

Akkio has introduced Agency Data LLM, the first Large Language Model tuned to advertising analytics. 

9. New medical LLM, PathChat 2, can talk to pathologists about tumors, offer diagnoses

Four state-of-the-art large language models (LLMs) are presented with an image of what looks like a mauve-colored rock. Read More 

10. Skymizer launches ET2 IP solution, creating more possibilities for LLM with hardware and software platform 

After announcing its foray into the LLM (Large Language Model) IP market, Skymizer recently unveiled a series of hardware and software solutions centered around LLM IP. Read More 

11. Nvidia Unveils Open Model That Creates LLM Training Data

As concerns grow that large language models (LLMs) are running out of high-quality training data, Nvidia has released Nemotron-4 340B, a family of open models designed to generate synthetic data for training LLMs across various industries. Read More 

Upgrades

1. Google translate to add support for 110 new languages through Palm 2 LLM 

Google has been aggressively using AI capabilities to transform functionalities and expand on features that were limited before the technology came about. Now, the company is using its Large Language Model to add 110 new languages to Google Translate, almost double the number of languages offered previously. Read More 

2. Huawei Cloud readies Pangu LLM to meet possible demand in SA

Huawei Cloud confirmed its Pangu large language model (LLM) could be made available to South Africa should this market show demand for the technology as the company moves to entrench its AI-driven Everything-as-a-Service (EaaS) strategy. Read More 

3. NTT's 'tsuzumi’ LLM is now available on Microsoft Azure 

NTT Data will start offering its large-scale language model (LLM) tsuzumi on Microsoft's model-as-a-service (MaaS) platform, which makes LLMs available in the cloud, from November 2024 in Japan. Read More 

4. Gradient creates an open LLM with a million-token context window

In a recent collaboration, AI startup Gradient and cloud compute platform Crusoe extended the “context window” of Llama-3 models to 1 million tokens. Read More 

Research & Technology News

1. Lower energy, high performance LLM on FPGA without matrix multiplication:

A new technical paper titled “Scalable MatMul-free Language Modeling” was published by UC Santa Cruz, Soochow University, UC Davis, and LuxiTech. Read More 

2. Researchers run high-performing large language model on the energy needed to power a lightbulb:

UC Santa Cruz researchers found that they could power a billion-parameter-scale language model on just 13 watts, about equal to the energy of powering a lightbulb. Illustration by Molly Fine. Read More 

3. Ensuring Integrity: Secure LLM  tokenizers against potential threats:  

In a recent blog post, NVIDIA's AI Red Team has shed light on potential vulnerabilities in large language model (LLM) tokenizers and has provided strategies to mitigate these risks. Tokenizers, which convert input strings into token IDs for LLM processing, can be a critical point of failure if not properly secured, according to the NVIDIA Technical Blog. Read More 

4. Digital Transformation in higher education: The impact of LLM-powered chatbots:

Researchers Achraf Hsain and Hamza El Housni from Al Akhawayn University, Morrocco explore integrating GPT-3.5 and GPT-4 Turbo-powered chatbots into higher education to enhance internationalization and digital transformation. Read More 

5. Alibaba Cloud reveals its datacenter design, homebrew network used for LLM training:

Alibaba Cloud has revealed the design of an Ethernet-based network it created specifically to carry traffic for training large language models – and has used in production for eight months. Read More 

In the News 

1. LLM  price war sees cream rise to top:

The simmering price war among China's biggest artificial intelligence-model builders will greatly accelerate the launch of more advanced AI language models in the country to close the gap with US counterparts, said top executives from leading Chinese AI startups. Read More 

2. Alibaba’s large language model tops global ranking of AI developer platform Hugging Face:

Alibaba Group Holding’s family of large language models (LLMs) – the technology underpinning generative artificial intelligence (AI) services like ChatGPT – has been ranked among the world’s top 10 open-source models. Read More 

3. Why developing countries should look to China to leapfrog into LLM markets 

China's AI startups can't compete with tech giants on financial capital and computing power, so they must focus on AI applications in areas where they have unique data advantages. Read More 

4. Singtel, Deutsche Telekom, E&, SK Telecom, SoftBank sign JV agreement for telco LLM 

The founding partners of the Global Telco AI Alliance, Deutsche Telekom, SK Telecom, e&, Singtel and SoftBank have this week signed a joint venture for telco AI development. Read More 

5. Document crunch wins 2024 ai breakthrough award for "best overall large language model 

Document Crunch Wins 2024 AI Breakthrough Award for "Best Overall Large Language Model". Contract intelligence firm is only construction solution recognized, following in the footsteps of OpenAI's win in 2023. Read More