Sentence-BERT

A method for producing fixed-size sentence embeddingsembedding, n.: A list of numbers that stands for a word, picture or other item, arranged so that similar items get similar lists. It lets a computer measure similarity of meaning as distance. by fine-tuning a pretrained BERTBERT, prop. n.: An influential 2018 Google language model that learned to understand text by filling in hidden words. It made search engines and many language tools much better at grasping meaning in context. or RoBERTa encoder in a SiameseSiamese network, n.: A matching machine built from twin copies of one neural network: each copy turns one of two inputs, say two signatures or two face photos, into a summary, and the two summaries are compared. or triplet arrangement: each sentence passes separately through the same encoder, its token outputs are mean-pooled into one vector, and training pulls related sentences together, originally through a classifier over (u, v, |u − v|) on natural-language-inference pairs and a cosine similaritycosine similarity, n.: A score from −1 to 1 for how closely two lists of numbers point the same way, ignoring their size. It is the standard test of whether two texts mean similar things. regression on graded similarity scores. Because each sentence is encoded once and vectors are compared cheaply, the method removed the quadratic cost of running a cross-encodercross-encoder, n.: An AI model that reads a question and a candidate passage together, side by side, and outputs one relevance score. Accurate but slow, so it is used to reorder a short list. on every pair: the paper's example of finding the most similar pair among 10,000 sentences fell from about 65 hours to about 5 seconds. Its open-source sentence-transformers library became the standard toolkit for training and serving bi-encoderbi-encoder, n.: An AI design that turns a question and a document into number lists separately, so documents can be processed once in advance and compared with any new question in a flash. embedding modelsembedding model, n.: An AI model whose only job is to turn text (or images, or code) into lists of numbers that capture meaning, so a computer can measure how similar two things are. for semantic searchsemantic search, n.: Search that understands what you mean, not just the words you typed, so a search for 'cheap places to stay' also finds 'budget hotels'., clustering and retrieval-augmented generationretrieval-augmented generation, n.: Giving an AI an open-book exam: before answering, it looks up relevant passages in a set of documents and bases its reply on them, often citing where each point came from..

We embed the FAQ with an SBERT model and send the top five matches to the reranker.