{"version":"2.1.281","anchor":"precomputed-compact-arm-fraction-computed-from-a-new-facts","canonical_anchor":"precomputed-compact-arm-fraction-computed-from-a-new-facts","heading":"Precomputed compact arm fraction computed from a new 'facts' input","tier":"internal","area":"Compaction","url":"https:\/\/changelogs.core-directive.com\/v\/2.1.281\/e\/precomputed-compact-arm-fraction-computed-from-a-new-facts","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.281","markdown":"### Precomputed compact arm fraction computed from a new 'facts' input\n\nA precomputed-compaction value is now worked out from a new \"facts\" input instead of straight from the main model setting\n\n**Unclear.** The finding does not say what `facts` contains or whether this changes when compaction happens.\n\n**What**\n\nCompaction is when Claude Code summarises older parts of a conversation to free up room. For precomputed compaction, the arm fraction and the window source are now worked out from a new `facts` input instead of directly from `mainLoopModel`, the model the main conversation uses.\n\n**Why**\n\nThese values no longer depend only on which model is in use.\n\n- Area: Compaction\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5"}