Unverified50% confidenceOpinionExact time
RefCraft's low scores reveal that models struggle to extract and fuse relevant information from multiple reference screenshots, often relying excessively on text instructions or simply copying the references.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/20/2026
First Seen
Valid until: 1/18/2027
Sources
Related Entities
Related Claims
UnverifiedBad answers blamed on the model being inadequate mostly trace back to the documents themselves after investigation: outdated versions, duplicated content, and multiple files conflicting with each other.71% similarUnverifiedThe root cause of poor knowledge base performance typically lies in unreasonable document chunking strategies, inappropriate vector model selection, or insufficient retrieval recall quantity, rather than issues with the model itself.69% similarUnverifiedEven when the model is explicitly required to answer only based on retrieved documents, it still fills in gaps using parametric memory and generates seemingly well-formatted citations.69% similarUnverifiedRequiring the model in the system prompt to append Markdown footnotes with URLs is extremely unreliable in long-context or multi-document mixed scenarios, and the model will miss citations, cite incorrectly, or even fabricate sources.69% similarUnverified低质量的知识库会导致幻觉式引用，即模型看似使用知识库内容实则引用了错误或无关片段68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1073316API
curl https://kongchang.com/api/v1/knowledge/claims/1073316MCP
get_claim(id=1073316)