引導式測驗 · 中等 等級
自然語言推理與蘊涵測驗
了解 NLI 的蘊含、矛盾和中性標籤、基準資料集、變壓器輸入和證據檢查限制。
問題 1 的 8
自然語言推理的三個標準標籤是什麼?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- 自然語言推理的三個標準標籤是什麼?
- 蘊含、矛盾、中性
- 正、負、中性
- 真、假、未知
- 主詞、動詞、受詞
- 在 NLI 中,「蘊含」是什麼意思?
- 兩句話無關
- 給定前提,假設必須為真
- 給定前提,假設一定是錯的
- 這些句子在文法上是相同的
- 給定前提“一個人在舞台上彈吉他”,哪個假設是矛盾的?
- 一個人在演奏音樂
- 群眾都喜歡這首歌
- 舞台空無一人
- 音樂家拿著樂器
- NLI 資料集中的標註人工痕跡是什麼?
- A model that reasons perfectly with very little data
- A dataset where every annotator agrees
- A premise that is always longer than its hypothesis
- A spurious cue that lets a model guess labels without reasoning about the premise-hypothesis relation
- 哪些資料集被廣泛用於訓練和評估 NLI?
- SNLI 和 MultiNLI
- ImageNet 與 CIFAR
- MNIST 和 SVHN
- 維基文本和 Penn Treebank
- How can NLI support a check of a generated claim against source evidence?
- Measure the generated claim’s length
- Classify whether the evidence entails, contradicts, or leaves the claim neutral
- Count unique words only
- Translate the claim into another language
- In NLI, what does the neutral label mean?
- The hypothesis is guaranteed true
- The hypothesis is guaranteed false
- The premise neither entails nor contradicts the hypothesis
- The two sentences are grammatically identical
- How does a common transformer NLI classifier present a sentence pair to the model?
- Encode only the hypothesis and discard the premise
- Convert both sentences into audio first
- Train a separate model for every word
- Concatenate the premise and hypothesis with a separator token