Genpilot

Genpilot is an AI‑powered large language model that combines the knowledge base, literature database, tool library, database and online search function of BGI. It integrates intelligent question‑answering and writing functions to provide comprehensive solutions for DCS Cloud users and researchers.

• For DCS Cloud users: Genpilot not only uses the native model to solve basic questions but also supports online search to obtain the latest information for summarization and processing. In addition, genpilot embeds the operation manual of the DCS Cloud product and provides intelligent customer service.

• For researchers: Genpilot provides intelligent assistance for the entire academic research process. Through the Literature library, Intelligent writing and NFSC review, researchers can efficiently handle repetitive and time‑consuming research tasks, explore innovative inventions and discoveries, and achieve rapid transformation and writing of final research results.

Introduction to Large Language Model

Large Language Model (LLM) is an artificial intelligence model based on machine learning and natural language processing techniques. It can understand human language and has a certain level of memory capability, allowing it to generate contextually relevant answers. Although LLMs have made significant progress in text generation, there are still some challenges in terms of accuracy and handling rare questions. In the field of spatial‑temporal omics and professional life sciences, LLMs still face issues such as outdated information and a lack of specialized language resources.

Genpilot, building upon the powerful text processing capabilities of LLMs, enhances data construction, focuses on user experience, and improves the accuracy, professionalism, and timeliness of answers. However, Genpilot still faces some technical limitations that have yet to be overcome. The following are the capabilities and limitations of Genpilot:

• Capabilities:

1. Intelligent Q&A: Genpilot can understand user questions and make reasonable inferences and predictions based on data, providing relatively accurate answers.

2. Memory capability: Within the same conversation, Genpilot can remember users' historical questions and answers within a certain range, generating more coherent and logical responses.

3. Access to the latest information: Genpilot parses user questions and performs online searches, combining search results to generate the latest information and insights.

4. Access to professional information: Genpilot incorporates underlying data such as internal documents, literature, tools and knowledge bases, enabling access to more valuable and complex information.

• Limitations:

1. Length limitation: Due to the current limitations of large language models, Genpilot is restricted by token limits in user input, answer generation and memory capabilities, only considering limited context information.

2. Data bias: Large language models may still encounter situations where they cannot answer or provide incorrect answers, requiring users to evaluate and use them cautiously.

3. Single modality: Currently, Genpilot only supports text input and output.

Definition of terms

Below are some terms mentioned in the Genpilot User Manual, for the sake of avoiding ambiguity, I hereby provide explanations.

• Dialogue: It includes the user's question and the corresponding answer generated by Genpilot. A dialogue refers to a question and its answer, where there may be multiple answers.

• Chat: It refers to a collection of dialogues.

• Bot: A special tool designed to provide additional functionality and capabilities to Genpilot. By using a bot, Genpilot can access databases, APIs, etc., and integrate the obtained information to generate answers in order to better address various tasks and needs.

• Application: Refers to the "Intelligent Q&A" application provided by Genpilot for DCS Cloud users, as well as the applications provided for researchers: Literature library, Intelligent writing and NFSC review. Different applications correspond to different databases and functionalities.

• Token: It is the smallest unit of text in the large language model. It can be a word, a punctuation mark, a number, a symbol or other language elements. Each token has a corresponding encoding representation, and the model uses these encoding representations to process and generate text. For English, 1 token is approximately 4 characters or 0.75 words; for Chinese, 1 token is approximately 0.5 characters. Currently, there is a token limit for the large language model when processing and understanding text.

• Streaming Output: It refers to the gradual generation of results in a continuous stream, rather than generating all results at once. Genpilot returns answers in a streaming output format.

Intelligent Q&A

An application designed for DSC Cloud users, allowing them to choose from different bots such as the native model, online search, or DCS support service to obtain corresponding answers.

Before initiating a Q&A session, users can select different models. In-house model refers to the model built by DCS Cloud based on deepseek 32B. If the model selection is left empty, it defaults to the In-house model.

Native model

Freely converse and chat with the native model. Users can freely explore the capabilities of the native model through direct conversation.

Online search

Perform a web search on the user's question, summarize the search results to respond to the user's query, and provide access to the latest information, insights, and reference websites.

DCS support

Leverage the DSC Cloud knowledge base (including user manuals, documentation, FAQs, and other resources) to deliver more professional services and accurate troubleshooting for DSC Cloud-related inquiries.

Literature Library

Literature search

literature search can retrieve more than 0.1 billion articles from Pubmed and other sources, meeting the needs of different field preferences.

Intelligent subscription

currently provides two conditions of journals and IF, as well as keywords in the field of interest, and supports the literature subscription function of different journals and directions.

Recently read, my Document Library (default)

recently consulted documents are displayed in the recent reading module if no document library has been established, the collected documents will be displayed in my Document Library by default.

Sample Literature Library

We provide two official sample literature libraries, the single-cell spatio-temporal analysis tool Literature Library and the liver cancer spatio-temporal and single-cell Literature Library, for reference.

The New Document Library

can be added by clicking "+" after my document library, and the documents of personal interest can be uploaded freely for unified management and reference.

Papers Reading

Click on any document to pop up the document details, including the basic information of the document, the cited document, the generated citation and the AI speed function, which can quickly access the key information such as the title, author, abstract, journal, impact factor and so on.

• Click on the AI speed to quickly refine the question and answer points of the literature.

Single papers reading

click PDF AI full text interpretation can be a single document full text interpretation and intelligent question and answer.

• The positioning AI question-and-answer module can conduct intelligent question-and-answer exchanges on the contents of the literature and quickly obtain the information you need.

The AI summary helps refine the article tools and usage methods, refine and summarize the overall framework of the article: Introduction, methods, results and discussion parts, and help speed the literature.

Immersive Reading, free marking and key information annotation, easy to repeat the document and key information without forgetting.

Literature Library Multiple Literature Questions and Answers

Support the unified intelligent management of multiple documents in the new individual Literature Library, support intelligent question and answer and summary of multiple documents, and summarize the common research contents and methods of different documents in the literature library.

There are many questions and answers in the literature library. The answers are reasonable and well-founded. Specific paragraphs of specific documents are quoted to make the answers more accurate.

Intelligent writting

1.New paper

When creating a new paper, you must choose a template. Besides the blank template, Genpilot also provides two scientific research article templates:

Journal article: Includes Abstract, Introduction, Results, Method, Discussion

Research proposal: National Natural Science Foundation of China ‑ Young Scientists Fund ‑ Main Body Outline Template

In addition, by clicking "Outline by AI," users can also intelligently customize templates:

Enter information and design an outline:

View the generated outline:

Redesign: Modify the information to redesign the outline.

Regenerate: Generate an outline again based on the existing information.

Apply: Apply the outline to start writing the paper.

2.Edit

In addition to the regular document editing features, Genpilot offers intelligent editing functionalities:

• 4 ways to quickly bring up the AI intelligent editing tools:

1. Use the "CTRL+/" keyboard shortcut.

2. Click on "Genpilot" in the menu bar.

3. Select a piece of text, and choose the Genpilot editing tool from the hover tools:

4. Click on any function in the "SciWritting AI" on the right to quickly invoke that function:

• Intelligent Editing Function Description:

1. Inputting a question

Users can directly input a question and send it for inquiry

2. Content Editing

Intelligently edit all text in the article or text selected by the user:

• Continue Writing: Continue generating article content based on the original content.

• Summarize: Summarize the content.

• Translate: Currently only supports translation into Chinese/English.

• Improve Writing: Correct errors and enhance expression, etc.

• Fix spelling & grammar: Repair spell or grammatical errors.

• Change tone

  • Shorter: Use more refined language to describe the text.

  • Longer: Expand the text to make it more detailed and rich.

  • Professional: Change the description to align more closely with professional expression.

  • Scientific: Change the description to conform to objective, rigorous academic expression.

After generating content, the following actions can be done:

• Adjust: For the generated content, further improve writing, fix pinyin grammar, or switch style.

• Try again: Resend the request to generate content.

• Continue Writing: Continue generating content based on what has been already produced.

• Discard: Give up the content and close the intelligent editing tool.

• Copy: Copy the AI‑generated content and close the intelligent editing tool.

• Done/Replace: Insert the generated content into the document and replace the selected text.

3. SciWritting AI

Based on the user's input question or keywords, as well as all the text on the page or selected text, perform intelligent editing:

• Brainstorming: Generate multiple options to choose from.

The options generated are displayed under the "SciWritting AI " menu on the right and can be inserted into the article with one click:

• Generate content: Produce a draft of the article content, complete with references and citation information.

Content generated includes reference citation information, with references displayed under the "SciWritting AI " menu on the right:

• Title design: Generate article titles. The operation is similar to "Brainstorming".

• Introduction generation: Generate the introductory part of the article, complete with references and citation information. The operation is similar to "Generate content".

4）Output

Users can select the language of Genpilot's output and choose "Auto" to let the model determine the output language itself.

• Reference Management

1）Automatic generation of references

After the "Generate content" and "Introduction generation" content is inserted into the article, the references are automatically updated:

2）Add reference

Users can input bibliographic information to serve as references for the article:

3）Insert citation

Insert the citation of the references at the cursor position:

3.Preview and Download

Preview

Click the preview button to enter the article preview mode. In preview mode, users are unable to edit the article:

Download

Click the download button to choose downloading the article as html, pdf or word. The exported file will be named as the article title:

If you choose to download PDF, a print page will pop up. Choose "Save as PDF" as the printer option and click save. This will allow you to save the article as a PDF file locally:

NSFC Review

Provide professional national nature review function, click on the National Nature review module, upload the locally written National Nature book, click on the run, you can get professional review opinions.

Intelligent Q&A - Old

The Intelligent Q&A application has been fully upgraded. Data from the historical versions of the Intelligent Q&A application will be retained until March 31, 2025. Please promptly save your important data.

• The legacy version of Intelligent Q&A supports the following features:

1. View, export and delete historical conversations

2. View and delete user-created assistants in the Bot Center

• The following features are not supported:

1. AI Q&A, including initiating new conversations, continuing questions in historical conversations, etc.

2. Creating or publishing bots

3. Viewing historically uploaded literatures

4. Other data modification operations (excluding deletion)