Bidirectional Encoder
A model architecture represented by BERT that uses context from both the left and right sides when encoding each token. It is suited to discriminative tasks such as classification, intent recognition and semantic matching, typically with tens of millions to hundreds of millions of parameters.
Timeline (last 90 days)
A bidirectional encoder uses context from both the left and right sides when encoding each token, unlike GPT-style autoregressive models that only look at context from left to right.
Typical bidirectional encoders range in size from tens of millions to hundreds of millions of parameters, far smaller than generative models with billions of parameters.
Bidirectional encoder models can simultaneously see left and right context at every position, making them suitable for understanding tasks such as classification and entity recognition; both GLiNER and Jev are based on the bidirectional encoder approach.
All Facts (3)
A bidirectional encoder uses context from both the left and right sides when encoding each token, unlike GPT-style autoregressive models that only look at context from left to right.
50%UnverifiedTypical bidirectional encoders range in size from tens of millions to hundreds of millions of parameters, far smaller than generative models with billions of parameters.
50%UnverifiedBidirectional encoder models can simultaneously see left and right context at every position, making them suitable for understanding tasks such as classification and entity recognition; both GLiNER and Jev are based on the bidirectional encoder approach.
50%