Unverified50% confidenceFactExact time
Prefilling refers to pre-filling the beginning of the assistant's reply in a conversation to guide the model to continue in a particular direction.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe three-step method for starting a clean new conversation: 1. Have the model do a fact review in five bullet points (including only what the user has said); 2. Manually check the review content, especially the numbers; 3. Open a new conversation, paste the review at the beginning, and then make the new request.77% similarUnverifiedThe System Prompt is injected by developers before a conversation begins, declaring to the model its identity, behavioral boundaries, and reply style, and is placed at the very front of the context window in each conversation74% similarUnverifiedDuring multi-turn tool calls, each tool execution result is appended to the conversation history as a new tool_result message, and the model reads the outputs of previous steps from the conversation context rather than retaining intermediate values in internal memory.72% similarUnverifiedContinuing within the same session is more effective than starting a new session and pasting a report, because the former allows the model to directly access the complete reasoning chain.71% similarUnverifiedMid-conversation steering allows developers to inject new instructions or constraints during the model's generation of a response, enabling real-time intervention.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1101828API
curl https://kongchang.com/api/v1/knowledge/claims/1101828MCP
get_claim(id=1101828)