Large Language Models (LLMs) are a type of artificial intelligence that specializes in understanding and generating human-like text. These models are trained on extensive datasets, enabling them to respond to queries, write content, and engage in meaningful dialogues. In the realm of medicine and research, their ability to process and summarize vast amounts of information quickly makes them valuable tools for research and learning. If you want to learn more visit our Wiki page.

We recommend you start at the Wiki page, but here is a snapshot.

Large Language Models (LLMs) are versatile tools, akin to computers or statistical software, and their utility depends greatly on your specific needs. Here are some ways you can leverage them in radiation oncology:

• Summarizing Text: Quickly digest large volumes of text, such as research papers or medical guidelines.

• Content Generation: Create educational materials, reports, or even draft emails.

• Text-to-Code: Convert natural language into code for data analysis or automation tasks.

• Brainstorming: Generate ideas for research projects, clinical trials, or patient care strategies.

• Data Wrangling: Assist in organizing and analyzing complex datasets.

• Administrative Support: Automate routine tasks like scheduling and record-keeping.

The most exciting aspect of LLMs is the potential for new, innovative uses that we have yet to discover. We're eager to see how our community finds creative applications for these tools.

Important Disclosure: While LLMs offer numerous benefits, it's crucial to approach them with a critical eye. They are not infallible and can sometimes "hallucinate" or produce inaccurate information. Always exercise due diligence when using these tools. Additionally, it's important to remember that LLMs like ChatGPT are not HIPAA certified. Therefore, avoid uploading or sharing confidential or patient-specific information. For more information on best practices and responsible use, check our wiki.

• Select a Platform: Choose a platform that offers LLM access, like OpenAI's website for ChatGPT or Claude by Anthropic.

• Create an Account: If required, sign up with your email and password.

• Familiarize Yourself: Explore the interface, typically featuring a text input area for your queries.

• Start Simple: Begin with basic questions or "prompts" to understand how the LLM responds.

• Experiment: Gradually try more complex prompts to explore the model's capabilities.

• Review Responses: Always check the LLM's output for accuracy and refine your prompts as needed.

• Use Responsibly: Remember, LLMs are tools and should not replace professional judgment.

• Join Online Communities: Engage with user forums for tips and shared experiences, such as Reddit or Github.

• Stay Informed: Keep up with updates and new features in LLM technology.

If you want to learn more visit our Wiki page.

Click on a topic below and you will find a curated set of prompts that you can use with your favorite LLM. You can use those for inspiration and edit them accordingly. You will note for each prompt there is a sample response to show you how an LLM would respond to you.

Here is a list of exciting papers we are reading. There are new papers being published everyday, we are our adding our paperpile that you access here.