Unverified50% confidenceFactExact time
The research team proposed the Knowledge Transfer Probing (KTP) protocol, which injects new knowledge into a model through image generation training, then uses pure text question answering to test whether this knowledge has truly been learned by the model.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedEnterprises can use the outputs of closed-source models to build training datasets, and then transfer large model capabilities to private small models through knowledge distillation.69% similarUnverifiedNative large language models are trained on massive publicly available internet data and generate probabilistic responses based on learned knowledge.69% similarUnverifiedLarge model projects face engineering issues such as hallucinated data fabrication, outputting pseudo-tool tags, and using outdated training knowledge to override the latest data returned by tools.68% similarUnverifiedThis experiment demonstrates a replicable technical approach: a large model defines the style and generates training data, while a small model handles real-time deployment—a workflow with reference value for game filters, real-time video processing, AR, and similar scenarios.67% similarUnverified知识图谱以工具（tools）形式暴露给前沿模型，而非直接塞进 prompt，使模型可按需多次调用查询67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1046203API
curl https://kongchang.com/api/v1/knowledge/claims/1046203MCP
get_claim(id=1046203)