Idanwo ti o sopọ mọ itọsọna · Alabọde Ipele

Atunyẹwo Idanwo

Review how rerankers reorder a retrieved candidate set, why cross-encoders are slower, and how first-stage recall limits the final results.

Jẹmọ itọsọna awọn ọnaAtunṣe atunṣe
Ibeere 1 ti 8

Kini ipa ti oluṣatunṣe ninu opo gigun ti epo igbapada?

Awotẹlẹ gbogbo ibeere8 ibeere

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

  1. Kini ipa ti oluṣatunṣe ninu opo gigun ti epo igbapada?
    • Lati tun- Dimegilio oludije kekere ṣeto nitorina awọn abajade ti o yẹ julọ dide si oke
    • Lati ọlọjẹ gbogbo koposi lati ibere
    • Lati ṣe ipilẹṣẹ awọn iwe aṣẹ tuntun
    • Lati compress atọka wiwa
  2. Kini o ṣe iyatọ si koodu-agbelebu lati oluyipada-meji?
    • Ayipada koodu-agbelebu ṣe ifibọ ibeere ati iwe ni lọtọ
    • Ayipada koodu-agbelebu n ṣe ifunni ibeere ati iwe papọ nitorina akiyesi ṣe afiwe wọn ni apapọ
    • Ayipada-alakoso ko le lo awọn ayirapada
    • Ayipada-agbekọja nikan ṣiṣẹ lori awọn aworan
  3. Kilode ti a fi lo awọn koodu-agbelebu fun isọdọtun dipo igbapada ipele akọkọ?
    • Wọn ko pe
    • Wọn mu iwe kan nikan mu lailai
    • Wọn ti lọra pupọ lati ṣiṣe lori gbogbo kopu kan nitori pe awọn olutọpa iwe ko le ṣe iṣiro tẹlẹ
    • Wọn ko nilo iṣiro
  4. Kini olugbapada ipele akọkọ ṣe deede fun?
    • Itọkasi ni ipo 1 nikan
    • Ṣiṣẹda awọn idahun ede adayeba
    • Titunṣe awọn iwe aṣẹ marun gangan
    • ÌRÁNTÍ ati iyara kọja gbogbo koposi
  5. How can a reranker affect a RAG candidate set, and what can it not guarantee?
    • It can reorder retrieved candidates, but cannot add a missing passage or verify generated claims
    • It guarantees the most relevant source is present and every generated claim is correct
    • It replaces the first-stage retriever and searches the entire corpus
    • It makes retrieval precision and recall perfect
  6. Ewo ninu iwọnyi jẹ ọna isọdọtun ti n yọ jade?
    • Piparẹ olugbapada ati awọn idahun lafaimo
    • Lilo awọn LLMs bi awọn atunto atokọ ti o pinnu lori gbogbo eto oludije
    • Npadabọ si fifi aami si ọrọ afọwọṣe nikan
    • Rirọpo Ayirapada pẹlu ID igbelewọn
  7. Which metric can assess whether the first-stage retriever includes relevant passages before reranking?
    • Image resolution of the generated answer
    • BLEU score of an unrelated translation
    • Recall of relevant candidates in the retrieved set
    • Number of model parameters in the reranker
  8. Which metric did the Passage Re-ranking with BERT paper report a 27% relative improvement on for MS MARCO?
    • BLEU-4
    • ImageNet top-1 accuracy
    • Itumọ aṣiṣe pipe
    • MRR@10