Long Context

For 1M-token context decode, which parallelism axis actually relieves KV pressure?

Name the dominant cost driver of GPT-5.5 latency at a 200k-token input

Dividing pre-softmax attention scores by an extra factor > 1 at inference does what?

Explain effective context for Mistral's 4096 sliding window over a 16k input

Which token most often serves as an attention sink in pretrained autoregressive LLMs?

Order the three steps of a position-interpolation fine-tune that takes RoPE from 4k to 16k.

Llama-3 bumps RoPE base from 10k to 500k, which long-context property follows?

Name two production KV-cache eviction policies and what each prioritizes

In a 7B transformer, do attention or MLP layers eat more FLOPs, and does context length change the answer?

How are RoPE NTK scaling and YaRN used at inference time to extend context beyond the training length?

How much does FP8 KV cache help, what does it cost and what is KIVI?

Compute the KV cache size for a 70B model at 128k context (80 layers, 64 heads, head_dim 128, FP16).

Where does FlashAttention deliver the biggest serving speedup?

What's the 'attention sink' phenomenon, and how does StreamingLLM exploit it?

Match each sparse attention pattern to its design and exemplar architecture.

What problem does ring attention solve that data parallelism cannot?

Why is KV cache memory the bottleneck for long context LLM serving?

Match each long context strategy to what it modifies.