Verified80% confidenceFactTime unknown
Google disclosed in the Gemini technical report that its model was exposed to multimodal data during the pre-training phase, rather than first training a language model and then performing multimodal fine-tuning.
4
Sources
80%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAMIE's underlying architecture is based on Google's Gemini series of models, with specialized training and fine-tuning on medical literature, clinical guidelines, and simulated consultation data.73% similarUnverifiedGemini mixes text, image, audio, and video data as input from the pre-training stage to achieve native multimodal training.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/4711API
curl https://kongchang.com/api/v1/knowledge/claims/4711MCP
get_claim(id=4711)