Unverified50% confidenceFactExact time
For multimodal training, audio-vision models such as Gemma 3n and Gemma 4 can be fine-tuned using both audio and images in the same dataset.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/3/2026
First Seen
Valid until: 2/1/2027
Sources
Related Entities
Related Claims
Unverified多模态实现路径分为早融合（预训练阶段混合多模态数据）和晚融合（先用独立视觉编码器ViT提取图像特征，再通过投影层对齐到语言模型embedding空间）66% similarUnverifiedNative audio generation means the model directly outputs audio signals synchronized with the on-screen state while generating visual content.66% similarUnverifiedRADIO is a visual foundation model proposed by NVIDIA that generates visual representations balancing semantic alignment and fine-grained features by distilling knowledge from multiple vision-language pretrained models such as CLIP, DINOv2, and SAM.65% similarUnverifiedThe Minimax series of models supports synchronized audio-video generation, requiring the model to process visual frames and audio waveforms simultaneously within the same diffusion process.64% similarUnverifiedWOVEN is a data source and benchmark that annotates visual transition supervision along three dimensions—scene, action, and reasoning type—with data coming from rollouts of video pretrained generative models.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1131482API
curl https://kongchang.com/api/v1/knowledge/claims/1131482MCP
get_claim(id=1131482)