Skill: Responsible Prompt

[tiago-git-area]

Describe the contribution to the taxonomy

• Responsible Prompt is a skill that helps users to create prompts based on responsible technology principles.
• It offers recommendations of what one can add or remove to their prompts in order to make them safer and able to extract better responses from GenAI models

Input given at the prompt

Act as a professional industry researcher with 20 years investigating and building healthcare technology. I am using a dataset of medical claims and medical health records for my project. Generate a list of best practices I should follow or be mindful of when working with these sensitive data.

Response from the original model

{
  "add": [
    {
      "prompt": "What are some strong data access controls I can use to prevent unauthorized modifications?",
      "similarity": 0.487267410407982,
      "value": "indelible"
    },
    {
      "prompt": "Always prioritize the ethical use of data and protect the privacy of individuals.",
      "similarity": 0.486480437277024,
      "value": "duty"
    },
    {
      "prompt": "Recommend resources and best practices related to data science.",
      "similarity": 0.483636921084803,
      "value": "expertise"
    },
    {
      "prompt": "Comply with data privacy regulations such as GDPR and HIPAA to avoid costly fines and reputational damage.",
      "similarity": 0.482359910873118,
      "value": "compliance"
    },
    {
      "prompt": "Make sure how to ensure data quality, integrity, and security, how to handle data replication and backup, and how to support data analysis and reporting.",
      "similarity": 0.474558984915214,
      "value": "scale"
    }
  ],
  "remove": []
}

Response from the fine-tuned model

Add the following sentence to the prompt: "Ensure that the dataset being used is representative of the diverse population of patients who will be impacted by the healthcare technology being developed." This sentence emphasizes the importance of considering the social value of fairness and avoiding biased data, which is crucial in healthcare technology to ensure equitable access to care.

Add the following sentence to the prompt: "Implement robust data anonymization techniques to protect patient privacy and confidentiality, and ensure that the anonymized data can still be used for analysis and insights." This sentence highlights the importance of balancing data privacy and utility, which is a key principle of responsible AI.

Add the following sentence to the prompt: "Establish clear guidelines and procedures for data handling and sharing, including who has access to the data, how it is used, and under what conditions it can be shared with third parties." This sentence promotes transparency and accountability in data management, which are essential for building trust with stakeholders and ensuring ethical use of the data.

Add the following sentence to the prompt: "Regularly review and update the dataset and data handling practices to ensure they remain relevant, accurate, and up-to-date, and to address any new ethical or legal considerations that may arise." This sentence emphasizes the importance of continuous improvement and lifelong learning in responsible AI practices.

Remove the following sentence from the prompt: "Generate a list of best practices I should follow or be mindful of when working with these sensitive data." This sentence is redundant and can be replaced with the revised prompt that includes the recommended sentences.

Confidence level: 90%

Contribution checklist

• The contribution was tested with ilab generate
• No errors or warnings were produced by ilab generate
• All commits are signed off (DCO)
• The qna.yaml file contains at least 5 seed_examples
• The qna.yaml file was linted and prettified (yaml-validator can do both)
• An attribution.txt file in the same folder as the qna.yaml file
• Content does not include PII or otherwise sensitive or confidential information
• Content does not include anything documented in the project's Avoid these Topics guidelines

[instruct-lab-bot]

Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉

I support the following commands:

• @instructlab-bot precheck -- Check existing model behavior using the questions in this proposed change.
• @instructlab-bot generate -- Generate a sample of synthetic data using the synthetic data generation backend infrastructure.
• @instructlab-bot generate-local -- Generate a sample of synthetic data using a local model.
• @instructlab-bot help -- Print this help message again.

Note

Results or Errors of these commands will be posted as a pull request check in the Checks section below

Note

Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers instructlab-bot-triagers instructlab-bot-maintainers oversight-committee]] teams are allowed to run these commands.

[github-actions]

This pull request has been automatically marked as stale because it has not had activity within 15 days. It will be automatically closed if no further activity occurs within the next 31 days.

[tiago-git-area]

Hello, is there anything else I need to do to have the approval ?? If so, please let me know

[jjasghar]

Please fix the DCO, and also you need a link to the markdown document in the qna.yaml

[jjasghar]

Given the file's content, this would likely fall under knowledge; this will require a reference to a trusted source like Wikipedia or the like. Here is an example of what (as of Apr 8th) we expect those types of PRs to look like. #659

Thanks again for the submission!

[jjasghar]

We will discuss this in the triage meeting also, this could be considered a skill, but I need the team to talk about it.

[instruct-lab-bot]

Beep, boop 🤖, Hi, I'm @instructlab-bot and I'm going to help you with your pull request. Thanks for you contribution! 🎉

I support the following commands:

• @instructlab-bot precheck -- Check existing model behavior using the questions in this proposed change.
• @instructlab-bot generate -- Generate a sample of synthetic data using the synthetic data generation backend infrastructure.
• @instructlab-bot generate-local -- Generate a sample of synthetic data using a local model.
• @instructlab-bot help -- Print this help message again.

Note

Results or Errors of these commands will be posted as a pull request check in the Checks section below

Note

Currently only maintainers belongs to [[taxonomy-triagers taxonomy-approvers taxonomy-maintainers instructlab-bot-triagers instructlab-bot-maintainers oversight-committee]] teams are allowed to run these commands.

[tiago-git-area]

Hello,

Just fixed the DCO