ClaimMaster includes patent application drafting tools that let you store and invoke helpful OpenAI GPT prompts enriched with the text from your document to generate, paraphrase, and summarize text. GPT is an advanced large language model for natural language processing that can generate coherent and contextually relevant responses to given prompts, making it a very helpful tool for patent application drafting and summarizing Office Actions. ClaimMaster interfaces both with OpenAI and Microsoft Azure OpenAI GPT services (access to gtp-3.5-turbo model is currently free to +Drafting and +Shells subscribers) and also lets you pass your API account information for both services when completing GPT prompts (you’ll incur your own costs for the generated text). Some of the helpful tasks you can ask GPT to perform include:
- Draft detailed description of a figure – check out this post (GPT-4-Vision)
- Re-write a text portion to fix errors and make it more eloquent
- Add technical details to existing paragraph based on the discussed subject matter
- Explain or define a technical term or concept
- Write a paragraph or two explaining the state of problems with the current technology
- Generate a sample data table for a particular technical field
- Draft a section of the application based on the passed-in subject matter, such as title, background, or even claims.
In particular, ClaimMaster now includes GPT templates that allow you to use it for writing detailed descriptions of the provided patent figures. Note that this feature is only available for GPT vision models, so you’ll need to configure ClaimMaster to use the appropriate GPT model from OpenAI or Azure to enable this capability (e.g., gpt-4-vision-preview). For more information, check out this post.
To use the GPT tool in ClaimMaster, perform the following steps:
- Open the GPT Tool
From the ClaimMaster tab, click on the Patent Drafting Tools menu, then click on Generate Text with GPT menu:
- Select one of the text inputs from the document
ClaimMaster’s lets you enrich GPT prompts with the text from your document to generate, paraphrase, and summarize text for patent drafting. In the window that comes up, use the “inputs” tab to select one of the inputs from the document to enrich the stored GPT prompt template via the [INPUT] variable.
The following inputs are available: (1) text in the preview window – allows applying GPT transformations on the results produced by previous GPT requests, (2) selected text in the document, (3) surrounding word/sentence/paragraph at the cursor’s current position, (4) all document text, (5) part names extracted from the document, (6) acronyms extracted from the document, (7) terms extracted from claims, (8) claim text and individual limitations. - Select GPT Template and Preview the GPT Prompt (Not Sent Yet)
Select the desired GPT prompt template from the list of the available templates shown on the right side of the GPT window. You can filter the window to make searching easier if you have a lot of templates. For more information on how to configure GPT prompt templates for patent drafting, click here. You can also optimize GPT prompts using “prompt engineering.” Once you select the GPT template, ClaimMaster will generate a preview prompt in the preview window based on the selected template and [INPUT] text value you’ve selected in the previous step. You can make edits to the generated GPT prompt directly into the preview window.
If you have access to a GPT service with vision models, you can also utilize templates that allow you to use it for writing detailed descriptions of the provided patent figures. For more information on using GPT to draft figure descriptions, check out this post.
In addition, you can use the “randomness/creativity” slider to make GPT responses more or less random/deterministic. This setting corresponds to the temperature setting for the GPT. Higher values will make GPT output more random/creative, while lower values will make it more focused and deterministic. It’s a good practice to always review GPT responses for factual correctness, especially if you are using GPT prompts with settings for higher response randomness/creativity.
- Select Desired GPT Settings (Optional)
You can also optionally specify your preferred GPT service and other model information. If you have your own account with OpenAI or Azure OpenAI, ClaimMaster lets you pass your API account information for completing GPT prompts (you’ll incur your own costs for the generated text). You may find this option beneficial as you’ll be able to use your custom end-points (Azure) or more advanced GPT models (ClaimMaster currently uses gtp-3.5-turbo by default), as well as set the max limit on the tokens to generate more text than allowed ClaimMaster’s default settings. For more information, click here.
- Send Your Prompt to GPT
Press the Send prompt to GPT button to send the final prompt to GPT. Note that ClaimMaster will only transmit data to GPT after your explicit command to do so. You control exactly what data is being sent to GPT via the prompt, so you should review it before sending the prompt to GPT. Please note that you are responsible for determining whether OpenAI API policies and/or Azure OpenAI API policies suit your needs, satisfy your organization’s security requirements, and assume risk when using GPT services with your confidential data. Here’s more information.
- Review and Insert the Generated Text into the Document
After some processing, GPT service you’ve configured will return a response to your patent drafting prompt. It’s a good idea to review GPT responses for factual correctness, especially if you are using GPT prompts with settings for higher response randomness/creativity. You can make edits to text directly in the preview window. Once you are satisfied with the preview/edits of the text generated by GPT, proceed to insert the generated text into the document. You’ll do this by clicking the Insert text into document button. You can also choose to at the location of pre-configured background/title/abstract replacement field.
.
For more information on GPT “prompt engineering” in ClaimMaster, click here.
For more information on this feature, check out the Online Manual.