引導式測驗 · 硬 等級
最小貝葉斯風險解碼測驗
測試 MBR 解碼的預期效用、取樣偽參考、二次候選比較、度量限制和候選池約束。
問題 1 的 8
MBR 解碼選擇什麼而不是單一最可能的序列?
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。
- MBR 解碼選擇什麼而不是單一最可能的序列?
- 相對於其他可能的產出具有最高預期效用的候選人
- 最短序列
- 隨機樣本
- 最長序列
- 既然無法獲得真正的引用,MBR 使用什麼作為偽引用呢?
- 固定的外部語料庫
- 樣本候選池輸出本身
- 人工註釋
- 訓練集
- 為什麼樸素 MBR 的運算成本很高?
- 需要重新訓練模型
- 需要更大的詞彙量
- N 個候選者的成對比較是 O(N 平方) 度量調用
- 它不使用 GPU
- What has COMET been used for in minimum Bayes risk machine-translation experiments?
- As a universal guarantee of better translation quality
- As a tokenizer for the candidate pool
- As an exact reference translation available at inference
- As a learned utility metric in evaluated translation settings
- 推理的自洽性與 MBR 相關,因為它:
- 對許多連鎖店進行抽樣並選擇最受認可的答案
- 始終採用第一個採樣答案
- 內部使用集束搜尋
- 忽略模型的分佈
- How is risk generally expressed in minimum Bayes risk decision-making?
- Always one minus a normalized similarity score
- Expected loss under a chosen loss function and reference distribution
- The model’s parameter count divided by accuracy
- The number of candidates sampled
- If the chosen utility metric disagrees with human judgments, what will MBR primarily optimize?
- Human preference automatically overrides the metric
- The model’s raw next-token likelihood
- The expected utility under that selected metric
- The shortest candidate in the pool
- If a high-quality answer is missing from the sampled candidate pool, what can MBR do?
- It guarantees the missing answer is generated
- It changes the model weights to add the answer
- It replaces the utility metric with a human reference
- It can only select or rerank candidates that are in the pool