Cross-Encoders vs Bi-Encoders Kediska

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Waa maxay faraqa qaabdhismeedka qeexaya ee u dhexeeya cod-beereye iyo cod-bixiye?
   • Qalab iskutallaab ah ayaa u habeeya labada qoraalba si ay calaamaduhu uga wada qaybgalaan, halka laba-kumbuyuutarku uu si gooni ah u kood u koodinayo qoraal kasta
   • Bi-encoder-ku wuxuu isticmaalaa fiiro gaar ah oo iskutallaabtu ma isticmaasho feejignaan
   • Qalabka-kumbuyuutarku wuu ka yar yahay oo ka dhaqso badan yahay codeeyaha laba-geesoodka ah
   • Labada-encoder-ku wuxuu xamili karaa hal qoraal oo keliya markiiba
2. Waa maxay sababta bi-encoders ay u cabbiraan malaayiin dukumeenti si hufan?
   • Waxay isticmaalaan xogta tababarka yar
   • Ku-xidhka dukumeentiyadu way madax-bannaan yihiin oo waa la xisaabin karaa oo la kaydin karaa
   • Weligood ma isticmaalaan transformers
   • Waxay ka boodaan tillaabada calaamadaynta gaabis ah
3. Dhuumaha wax-soo-saarka caadiga ah ee raadinta, sidee labada qaab-dhismeed la isugu daray?
   • Qalab iskutallaabtu waxa ay soo saartaa musharixiinta, ka bacdi laba-koodeeyaha ayaa dib u darajo ku sameeyaa iyaga
   • Labaduba waa barbar socdaan, codadkana waa la isku celceliyey
   • Labada-encoder-ku waxa uu soo celiyaa musharraxiin badan, ka bacdi encoder-ku-gudbiyeedka ayaa dib u dhigaya kuwa sare
   • Mid kaliya ayaa waligeed la isticmaalaa, marna labadaba
4. How does a late-interaction model such as ColBERT combine token-level query and document representations at scoring time?
   • It averages both texts into a single vector before any interaction
   • It concatenates raw audio waveforms and applies a vocoder
   • It ranks documents only by their character count
   • It compares query-token vectors with document-token vectors using a MaxSim-style aggregation
5. Muxuu sida caadiga ah u soo saaraa laba-encoder-ka hal qoraal?
   • Hal-dherer go'an vector dhexda
   • Dhibcaha khuseeya ee u dhexeeya laba qoraal
   • Liiska dukumentiyada
   • Qaybinta itimaalka ee erayada
6. After jointly processing a query and candidate, what does a cross-encoder produce?
   • A reusable document vector computed independently of the query
   • A relevance score for that specific query-candidate pair
   • A separate token vector for every item in the full corpus
   • A guarantee that the top result is correct
7. What kind of representation does a late-interaction model such as ColBERT retain for a text?
   • Only one global label for the whole corpus
   • A decoded audio waveform
   • A vector representation for each token
   • A complete copy of the cross-encoder’s pairwise scores
8. What similarity function do Sentence Transformers examples commonly use for comparing normalized bi-encoder embeddings?
   • Edit distance over raw audio samples
   • A pixel-wise image loss
   • A speech-recognition word error rate
   • Isku ekaanshaha cosine