M42 Introduces Med42: An Open Access Clinical Large Language Model (LLM) to Expand Access to Medical Knowledge

M42 Health, based in Abu Dhabi, United Arab Emirates, has just published Med42, a promising new open access big language clinical model. The release of this 70 billion-parameter model is a watershed moment in the effort to increase public access to advanced ai capabilities that can revolutionize healthcare.

Med42, refined from Meta’s Llama-2 – 70B model, surpasses its predecessors in open source medical ai by a wide margin. The model outperforms OpenAI’s ChatGPT 3.5 on many medical question answering datasets, achieving up to 72% accuracy in a zero-shot assessment on the USMLE. This demonstrates Med42’s ability to aid clinical decision making by providing physicians with easy access to synthesized medical knowledge.

The M42 Health ai team built Med42 using its massive human-curated medical literature and patient information dataset. M42, Cerebras and Core42 (a subsidiary of M42) worked together to perfect the Condor Galaxy 1 supercomputer. The effectiveness of the model was also evaluated by experts from the Mohamed bin Zayed University of artificial intelligence (MBZUAI).

Med42 by M42 is a free, publicly available clinical large language model (LLM) created to make more medical information open to the public. Based on LLaMA-2 and with 70 billion parameters, this generative ai system offers precise answers to medical queries.

One of Med42’s strongest points is its adaptability. As an aid to ai, it has the potential to significantly alter medical judgment. It can be used for everything from generating personalized treatment plans based on medical records to speeding up the process of searching through mountains of medical supplies.

As an ai aid with the potential to improve clinical decision-making and expand access to LLM for healthcare use, Med42 is now available for testing and evaluation. Examples of possible applications are:

• Answer health-related questions
• Synopsis of medical history
• In support of medical diagnosis.
• Common health questions

Med42 code and weights have been published on Hugging Face, encouraging a wide range of analysis and scientific input to foster collaboration and continued growth. Med42’s licensing terms are based on those of Meta’s Llama 2 model, making it available for free research and non-commercial use, but imposing appropriate restrictions to take into account the risks and liabilities associated with the use of ai in the medical attention.

Key performance indicators:

• Med42 outperforms the competition with 72% accuracy on a sample USMLE exam compared to other publicly available medical LLMs.
• The MedQA dataset results in an accuracy of 61.5% (GPT-3.5 is 50%).
• The clinical problem results of MMLU are consistently better than those of GPT-3.5.

Limitations:

• The therapeutic application of Med42 is still in its early stages. Extensive human testing is currently underway to ensure safety.
• The risk of creating misleading or dangerous data.
• Possible danger of using biased data for training.

Although the findings are encouraging, the researchers caution that further real-world validation of Med42 is necessary before it can be used in clinical practice. Problems can arise by producing inaccurate or harmful results or by failing to address existing biases in the training data. As Med42 moves beyond baseline values ​​and toward potentially substantial benefits for patients, M42 emphasizes the importance of responsible testing.

Med42 showcases the remarkable development of medical ai while emphasizing the importance of ethics and safety in research and development. Thanks to this, researchers around the world will be able to benefit from its open access publication. Models like Med42 can improve healthcare decision-making and expand treatment access on a global scale if they undergo extensive validation. Its launch is an important step forward in healthcare ai, but realizing its full potential will require continued openness and teamwork.

Review the Project page. All credit for this research goes to the researchers of this project. Also, don’t forget to join. our 31k+ ML SubReddit, Facebook community of more than 40,000 people, Discord Channel, and Electronic newsletterwhere we share the latest news on ai research, interesting ai projects and more.

If you like our work, you’ll love our newsletter.

We are also on WhatsApp. Join our ai channel on Whatsapp.

<!– ai CONTENT END 2 –>