引導式測驗 · 中等 等級

Logit 透鏡與中間層解碼測驗

測試中間殘差讀數、歸一化和非嵌入、調諧仿射探針、詞彙 Logits、softmax 和因果解釋的限制。

問題 1 的 8

What readout does a plain logit lens reuse to obtain vocabulary scores from an intermediate state?

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。

  1. What readout does a plain logit lens reuse to obtain vocabulary scores from an intermediate state?
    • A required new image classifier for every layer
    • The model's appropriate final normalization and unembedding output projection
    • A random dictionary sort with no model parameters
    • A separate retrained tokenizer for each prompt
  2. Why can the same output projection be applied mathematically at multiple residual-stream depths?
    • Every layer must contain an identical copy of the final state
    • The entire transformer is a single linear mapping
    • All intermediate token scores are guaranteed identical
    • The states have compatible hidden dimensions, although their representations may not be equally calibrated for that readout
  3. 相對於普通 Logit 鏡頭,「調諧鏡頭」解決了什麼問題?
    • 它透過學習的仿射變換來糾正每層的不匹配
    • 它消除了任何非嵌入的需要
    • 它使模型的推理速度更快
    • 它用 argmax 取代了 softmax
  4. If an intermediate lens readout favors a correct factual token, what has been directly observed?
    • The unique causal location of the model's factual knowledge
    • A guarantee that all later output will be correct
    • That the particular probe maps that state to a high score for the token
    • The exact reasoning process of a conscious system
  5. Logit 透鏡最直接的用途是什麼?
    • 壓縮模型權重
    • 語言之間的翻譯
    • 加密模型已激活
    • 觀察深度預測如何以及在何處形成
  6. Why can a plain final-head readout be misleading on early transformer states?
    • Early states always use a different vocabulary file
    • Their representation distributions can differ from the final state for which the head was trained
    • Every early-layer vector is necessarily unstructured noise
    • No matrix operation can be applied before the final layer
  7. A lens uses a model with a fifty-thousand-entry vocabulary and a 768-dimensional hidden state. How many vocabulary logits does its unembedding produce?
    • 768, one score per hidden coordinate only
    • One, because the token must be chosen before projection
    • Fifty thousand, one score per vocabulary entry
    • Exactly the number of transformer layers
  8. What does softmax add to a logit-lens vocabulary-score display?
    • A normalized distribution over the projected vocabulary scores
    • Proof that every displayed probability is a calibrated belief
    • A causal intervention on the model's hidden states
    • A replacement for the output projection's vocabulary dimension