Text Embedding
A technique that maps discrete linguistic symbols into a continuous vector space. Pretrained models encode text into fixed-dimensional dense vectors so that semantically similar texts are closer together in the vector space, forming the technical basis of semantic search and RAG systems.
Timeline (last 90 days)
Modern embedding models are usually based on the Transformer architecture (such as BERT and its variants), encoding text into fixed-dimensional dense vectors (common dimensions are 768, 1024 or higher).
Value representations should be extracted preferentially from model activations rather than relying on the text embeddings of value names.
MiniMax H3 supports both LoRA (low-rank adaptation) and text embeddings as methods for achieving character control.
Text embeddings encode a character's visual features as high-dimensional vectors injected directly into the model, requiring no additional training and enabling faster inference, but placing higher demands on the quality and quantity of reference images.
In terms of character consistency, LoRA is generally more stable but requires pre-training, while text embeddings are more flexible and better suited for on-the-fly character-swapping needs.
All Facts (5)
Modern embedding models are usually based on the Transformer architecture (such as BERT and its variants), encoding text into fixed-dimensional dense vectors (common dimensions are 768, 1024 or higher).
50%UnverifiedValue representations should be extracted preferentially from model activations rather than relying on the text embeddings of value names.
50%UnverifiedMiniMax H3 supports both LoRA (low-rank adaptation) and text embeddings as methods for achieving character control.
50%UnverifiedText embeddings encode a character's visual features as high-dimensional vectors injected directly into the model, requiring no additional training and enabling faster inference, but placing higher demands on the quality and quantity of reference images.
50%UnverifiedIn terms of character consistency, LoRA is generally more stable but requires pre-training, while text embeddings are more flexible and better suited for on-the-fly character-swapping needs.
50%