SA260324R-0293_r1

SA260324R-0293

① SA Source

• Source: 開啟完整 SA 文章
• Section: “
• Line hint: 374

Context Before

Evidence

KV cache grows linearly with input sequence length and number of users and is the primary tradeoff when it comes to prefill performance (§3§).

Context After

② Atomic Claim

KV cache 會隨輸入 sequence length 與 user 數量線性成長，並成為 prefill performance 的主要 tradeoff。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/KV cache",
        "label": "KV cache"
      },
      "role": "cache"
    },
    {
      "node": {
        "id": "04_knowledge_base/Sequence length",
        "label": "sequence length"
      },
      "role": "scaling_variable"
    },
    {
      "node": {
        "id": "04_knowledge_base/Prefill",
        "label": "prefill"
      },
      "role": "performance_context"
    }
  ],
  "qualifiers": {
    "condition_text": "KV cache also scales linearly with number of users",
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "relation_type": "KV_CACHE_SCALING_AND_PREFILL_TRADEOFF"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
cache	KV cache	04_knowledge_base/KV cache
scaling_variable	sequence length	04_knowledge_base/Sequence length
performance_context	prefill	Prefill

⑤ Human Review

請在 Properties 逐項確認：

• 原文 → Atomic Claim 是否忠實
• Atomic Claim → Semantic Frame 是否忠實
• Canonical Entity mapping 是否正確
• Epistemic mode 是否保留原文語氣
• 最後選擇 review_action

Review state

Markdown 內文不是正式 approval。只有 Apply bridge 寫入的 Decision Ledger event 才是正式決策。