DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression

Automated excerpt

The subsequent 5 layers of Reuse mode CSA2 will all reuse the Main KV and TopK indices produced by the first layer. Local KV (SWA KV): the sliding window cache of the most recent tokens (window_kv_cache). The main attention of one layer of CSA2 requires four tensors: Main Q, Main KV content, Top-K selection set, and local SWA KV.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.