Útmutatóhoz kapcsolódó kvíz · Közepes Szint
Multi-Token Prediction Képzési Kvíz
Review multi-token losses and shared-trunk heads, with DeepSeek-V3’s distinct sequential design, optional inference heads, and target offsets.
kérdés 1 -ból 8
In the cited Meta study, what architecture predicts several future tokens from a training position?
Tekintse meg az összes kérdést8 kérdéseket
A válaszok és magyarázatok megjelennek a kvíz kitöltésekor.
- In the cited Meta study, what architecture predicts several future tokens from a training position?
- A mandatory separate complete language model for every word
- Independent output heads above a shared transformer trunk
- One head that can target only the preceding token
- A larger vocabulary with no additional prediction targets
- Melyik modell használt különösen több token előrejelzési célt az előképzésben?
- Az eredeti Word2Vec
- Konvolúciós képosztályozó
- Egy n-gramos modell
- DeepSeek-V3
- What additional supervision does multi-token training supply at a prefix position?
- Loss terms for multiple future-token targets
- Only a larger batch size with the same single target
- A mandatory image-reconstruction target
- No prediction loss at any position
- How are auxiliary predictions used in the Meta paper’s self-speculative inference experiment?
- They are accepted unconditionally without verification
- They automatically retrain the model on every prompt
- They propose tokens that are then checked by the decoding procedure
- They encrypt every generated token
- Mi történik a segédfejekkel edzés után, ha nincs szükséged gyorsításokra?
- Maradniuk kell, és meg kell duplázniuk a modell méretét
- Eldobhatók, így a telepítési méret nem változik
- Figyelemrétegekké alakulnak
- Ők lesznek a tokenizátorok
- Which evaluated task family showed notable gains in the Meta multi-token study?
- Code generation benchmarks
- A guarantee of perfect code compilation in every language
- Only alphabetizing raw filenames
- A required quantum-sensor prediction task
- In a four-head offset design predicting from prefix position t, which token is the fourth head’s target?
- The token at t−4
- The token at t+1 for every head
- Only the token at t itself
- The token at t+4
- What distinguishes DeepSeek-V3’s stated MTP architecture from the cited independent parallel-head implementation?
- It predicts only tokens before the prefix
- It requires every target to be an image pixel
- Sequential auxiliary modules that preserve a causal chain across prediction depths
- It removes the language-model output vocabulary