ColBERT ati Olona-Vector Retrieval Quiz

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Bawo ni ColBERT ṣe aṣoju iwe-ipamọ kan?
   • Bi ọpọlọpọ fun-àmi contextual fekito
   • Bi awọn kan nikan ipon fekito
   • Bi apo fọnka ti awọn iṣiro ọrọ-ọrọ
   • Bi okun ọrọ aise nikan
2. Iṣe wo ni ColBERT nlo lati ṣe Dimegilio ibaramu ibeere-iwe?
   • Cosine ti meji nikan fekito
   • MaxSim: apao ti aami ibeere kọọkan baramu to dara julọ si eyikeyi ami iwe
   • Cross-entropy lori gbogbo fokabulari
   • BM25 igba igbohunsafẹfẹ
3. Kilode ti ColBERT ṣe yara ju koodu-agbelebu ni iwọn?
   • O nlo awọn fokabulari kekere
   • O kọju ibeere naa patapata
   • Awọn ifibọ àmi iwe ti wa ni iṣaju iṣiro ati atọka offline
   • O ṣe ilana nikan ọrọ akọkọ ti iwe-ipamọ kọọkan
4. Kini iṣoro pẹlu awọn atunpada ipon-ẹyọkan ni ColBERT koju?
   • Wọn lọra pupọ lati atọka
   • Wọn ko le lo awọn nẹtiwọọki nkankikan
   • Wọn ko nilo data ikẹkọ
   • Wọn padanu didara-dara, awọn alaye ipele-ami nipasẹ titẹkuro si fekito kan
5. Ilọsiwaju bọtini wo ni ColBERTV2 ṣafihan?
   • Iyọkuro ti o ku ti o dinku iwọn atọka pupọ
   • O yọ BERT kuro patapata
   • O yipada si fifi koodu-fekito kan
   • O ṣiṣẹ nikan lori awọn aworan
6. Kini ColPali lo ibaraenisọrọ pẹ ara ara ColBERT si?
   • Awọn ọna igbi ohun
   • Awọn abulẹ aworan ti awọn oju-iwe iwe, bii PDFs
   • Awọn koko-ọrọ ẹyọkan nikan
   • Database tabili awọn ori ila
7. In a two-stage search system, what role can ColBERT play after a fast first-stage retriever?
   • Generate the source documents from scratch
   • Replace the query with a document title
   • Rerank the candidate passages using contextual token-level matching
   • Discard every candidate before scoring
8. What does ColBERTv2's residual compression store for a token vector?
   • Only the raw document text with no vector data
   • A full copy of the language model for each token
   • One shared vector for the entire document
   • A centroid code together with a compact residual representation