Jamba Hybrid Transformer-Mamba Models Quiz

Während Sie am Quiz teilnehmen, werden Antworten und Erklärungen angezeigt.

1. Welche zwei Kernschichttypen verschachtelt Jamba?
   • Faltungs- und Pooling-Schichten
   • RNN- und CNN-Schichten
   • Transformer-Aufmerksamkeit und Mamba-Zustandsraumschichten
   • Diffusions- und Aufmerksamkeitsebenen
2. Welche zusätzliche Technik nutzt Jamba, um die Kapazität zu erhöhen und gleichzeitig die aktiven Parameter niedrig zu halten?
   • Expertenmix (MoE)
   • Wissensdestillation
   • Quantisierungsbewusstes Training
   • Verstärkendes Lernen aus menschlichem Feedback
3. How does a Mamba layer's recurrent representation differ from an attention layer's KV cache?
   • It removes every model parameter from memory
   • It skips all tokens beyond the first one
   • It stores an independent full copy of every earlier attention matrix
   • It maintains a fixed-size recurrent state instead of retaining a growing set of past keys and values
4. Welches Kontextfenster unterstützte das erste Jamba-Modell?
   • 4K-Token
   • 256.000 Token
   • 32.000 Token
   • 8 Millionen Token
5. What is a reason to retain attention components alongside state-space layers in a Jamba-style hybrid?
   • To combine different access and representation mechanisms while evaluating their quality-efficiency tradeoff
   • To guarantee perfect retrieval of every token on every task
   • To make all attention computation independent of sequence length
   • To eliminate the need for any model evaluation
6. How should Jamba's reported long-context memory benefit be evaluated for deployment?
   • Assume every 256K request fits any single GPU
   • Count only the recurrent state while ignoring weights and attention cache
   • Measure the actual model, precision, batch size, context, and hardware configuration
   • Treat every hybrid model as having zero sequence-state cost
7. Does original Jamba's use of many Mamba layers make the whole model's sequence memory constant as input length grows?
   • Yes; every hybrid attention cache has a fixed one-token size
   • No; its retained attention layers still store a growing KV cache
   • Yes; model weights disappear once the sequence becomes long
   • Only the tokenizer's file size changes
8. Why are total MoE parameters and active parameters per token different quantities in Jamba?
   • Every stored expert is necessarily deleted when inactive
   • A token must run all experts regardless of routing
   • The model's supported context directly sets its vocabulary size
   • Routing can activate only a subset of stored experts for each token