Guide-yakabatana mibvunzo · Pakati Nhanho

Mutongo-BERT Embeddings Quiz

Edza kuti mitsara inomisikidzwa inotsigira sei scalable semantic kutsvaga, kusanganisira Matryoshka anomiririra uye muchinjika-encoder reranking.

Nzira dzekutungamira dzinoenderanaMutongo Bert Embeddings
Mubvunzo 1 ye 8

Nderipi dambudziko guru rine plain BERT rinogadziriswa Sentence-BERT?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Nderipi dambudziko guru rine plain BERT rinogadziriswa Sentence-BERT?
    • Kuenzanisa mitsara yakawanda paviri neBERT kunonoka
    • BERT haikwanise kuverenga mutsara unopfuura mumwe
    • BERT inoshanda muChirungu chete
    • BERT haigone kugadziridzwa zvakanaka
  2. Ndeipi network yekuvaka iyo Sentence-BERT inoshandisa?
    • Imwe autoregressive decoder
    • A siamese network ine huremu hwekugovanisa shongwe dzeBERT
    • A convolutional image network
    • Mushandi wekusimbisa-kudzidza
  3. Which pooling strategy did the original Sentence-BERT paper evaluate for creating sentence embeddings?
    • Max achibatana pamusoro pemisoro yekutarisisa
    • Kushandisa hunhu hwekutanga chete
    • Zvinoreva kubatanidza pamusoro pematokeni ekumisikidza
    • Kusarudzika kusarudzwa kwechiratidzo chimwe
  4. Kana mitsara yaiswa, kufanana kwayo kunowanzoyerwa sei?
    • Kumhanya BERT zvakare pane ese ari maviri
    • Kuverenga mavara akagovaniswa
    • Kuenzanisa saizi yefaira
    • Cosine kufanana (dot product) pakati pevectors
  5. What training data did the original Sentence-BERT paper use for NLI fine-tuning?
    • Natural Language Inference examples labeled for entailment, contradiction and neutral relations
    • Only satellite images with land-cover labels
    • Only audio recordings with speaker labels
    • Only source code and bug reports
  6. In a possible semantic-retrieval workflow, what can a Sentence-BERT bi-encoder do?
    • Cross-attend to every query-candidate pair before any encoding
    • Encode a query and candidate texts separately so vector similarity can rank candidates
    • Guarantee retrieval quality for every corpus
    • Replace the need to evaluate relevance
  7. What does Matryoshka Representation Learning enable for an embedding?
    • It guarantees identical quality at every vector size
    • It converts embeddings into labeled text tokens
    • A single representation can be truncated to different dimensions to trade storage or compute for task performance
    • It removes the need for similarity comparisons
  8. In a common hybrid retrieval pipeline, what happens after a bi-encoder retrieves candidate passages?
    • The candidate text is converted into audio
    • The bi-encoder is discarded and no ranking occurs
    • The system averages all documents into one sentence
    • A cross-encoder can score query-passage pairs to rerank the candidates