Using the Chat Interface
Learn how to effectively communicate with AI models using the RoxWhy Chat interface
Table of Contents
Using the Chat Interface
Once you've selected an AI model, you can start chatting. The RoxWhy Chat interface is designed to be intuitive and powerful, allowing for natural conversations with AI models. This guide covers all aspects of using the chat interface effectively.
Chat Window Overview
The main chat window is where your conversation with the AI takes place:
- Chat Bubbles: Your messages appear aligned to the right (or labeled as "You"), while AI responses appear on the left (labeled with the model's name or "AI")
- Markdown Formatting: AI responses support rich formatting including bold text, lists, code blocks, and even LaTeX equations
- Scrollable History: Scroll up to view earlier parts of your conversation
Sending Messages
Interacting with the AI is straightforward:
1. Type your message or question in the Message Input Box at the bottom of the chat window 2. Press Enter or click the Send button (often shown as a paper plane icon) to send your message 3. Wait for the AI to generate and display its responseTip: If the platform supports "Ctrl+Enter to send" as a setting, you can enable or disable that in Settings. By default, pressing Enter will send the message.
Controlling AI Responses
You have several options to control how the AI responds:
- Stop Generation: While the AI is generating a long response, you can click the "Stop" button to halt the generation if you don't want to wait for it to finish
- Regenerate Response: After a response is complete, you may see a "Regenerate" option if you want the AI to try a different answer for the same prompt
- Context & Continuation: The conversation history provides context for the AI. It will consider the entire history of messages in that chat when formulating answers
Advanced Input Features
The chat interface supports several advanced input methods beyond basic text:
File Attachments
You can enhance conversations with document context:
1. Drag and drop files (such as PDFs or text documents) into the chat window 2. The system will load the document as a knowledge source for that conversation 3. Reference the file in your questions (e.g., "Based on the attached document, what is the summary?")
This feature is part of Retrieval Augmented Generation (RAG), allowing the AI to access and reference external documents.
Web Search Commands
If web search integration is enabled:
1. Start your message with a special command#
followed by URL or your normal query with 'Web Search Enabled'
Formatting and Code
Both you and the AI can share formatted content:
- Code Blocks: When sharing code, enclose it in backticks (\
\
\`) or use the "Code" button if available. The AI's code responses will appear in distinct blocks with syntax highlighting - Copy Button: Code blocks typically include a "copy" button for easily copying the code
- LaTeX Math: Mathematical expressions in LaTeX format will render properly in the chat
Best Practices
- Clear Questions: Keep your questions clear and specific for best results
- Follow-up Questions: Ask the AI to elaborate, simplify, or provide examples if an answer isn't clear
- New Conversations: For a new topic or a fresh start, begin a new chat so the context resets
- Experiment with Tools: Try the available tools (like uploading files or using search commands) to get more informative answers
- Formatting: Use formatting in your messages when appropriate, especially for code or structured information