Cuestionario vinculado a una guía · duro Nivel
Prueba de vectores de funciones y representaciones de tareas
Estudie cómo se extraen los vectores de funciones de las cabezas de atención, se inyectan en estados ocultos y se prueban en contextos de tareas.
Pregunta 1 de 8
What does a function vector represent in the studied language models?
Vista previa de cada pregunta8 preguntas
Las respuestas y explicaciones aparecen a medida que realiza el cuestionario.
- What does a function vector represent in the studied language models?
- A compact activation-space representation associated with a demonstrated task
- The model’s full weight matrix
- A single token embedding in every setting
- The optimizer’s learning rate
- ¿Cómo se construye normalmente un vector de función?
- Sumando todas las incrustaciones de palabras en el vocabulario.
- Al promediar los resultados de las cabezas de atención que realizan tareas específicas sobre indicaciones de ejemplo
- Ajustando el modelo en un ejemplo
- Leyendo las probabilidades finales de softmax
- What did the function-vector paper observe after injecting a vector into a zero-shot prompt?
- The model always crashed
- It guaranteed correct answers on any task
- It could trigger the associated task in the tested settings
- It permanently changed the model weights
- ¿Qué técnica de análisis sustenta la búsqueda de las cabezas que llevan un vector de función?
- Descenso de gradiente sobre la pérdida.
- búsqueda de haz
- Regularización del abandono
- Análisis de mediación causal.
- What do function-vector experiments provide evidence for?
- Some models form reusable internal representations associated with tasks
- All language models only memorize text
- Every model has one universal task vector
- Tokenization alone determines every behavior
- Why are function vectors interesting to steering research?
- They guarantee safe or correct behavior
- They provide an activation-level intervention that can influence task behavior in tested settings
- They eliminate prompts for every application
- They make all model decisions fully transparent
- At what part of the transformer did the paper report strong function-vector effects?
- Only the input embedding table
- Only the final output logits
- Middle layers in the tested models
- El vocabulario tokenizador
- What happened when researchers summed some function vectors?
- The vectors always canceled to zero
- The model’s weights were permanently updated
- Every possible task was composed perfectly
- Some combinations triggered new complex tasks in the tested experiments