Checkpoint the larger cap32 benchmark before results land

Preserve the new 32-track top-two benchmark entry point and current build-index phase so a later session can continue the stronger validation run without losing runtime context.

Constraint: The cap32 benchmark is still running, so only execution-state evidence is available
Rejected: Wait for cap32 results before recording anything | Risks losing the larger-benchmark checkpoint if the session ends first
Confidence: high
Scope-risk: narrow
Directive: Replace the cap32 running-state section with measured scores once hybrid eval.json and report.json land
Tested: Verified active cap32 processes; verified handoff records work-root, subset size, query cap, and current build-index phase
Not-tested: cap32 strategy scores because the run is still in progress

@@ -2,6 +2,28 @@
 
 ## 2026-06-02
 
+### Stage: 启动 cap32 top2 真实 FMA 对照并记录运行阶段
+
+完成项：
+- 启动更大的真实 FMA top2 benchmark：
+  - `work_root = /tmp/ab_smoke_seg_cap32_top2`
+  - `subset_size = 32`
+  - `max_test_queries = 20`
+  - 策略：`hybrid` vs `high_energy`
+- 更新 [session-handoff.md](./session-handoff.md)
+
+当前 fresh evidence：
+- `scripts/ab_smoke_segmentation.py ... --work-root /tmp/ab_smoke_seg_cap32_top2` 已启动
+- 当前 first lane 为：
+  - `hybrid`
+- 当前已进入：
+  - `run_demo.py build-index --resume --checkpoint-every-refs 100`
+- `report.json` 尚未落盘
+
+结论：
+- 现在已经开始验证 cap24 结论在更大 `subset=32` 上是否继续成立
+- 即使当前 session 结束，新 session 也可直接从 handoff 中的 cap32 入口继续盯结果
+
 ### Stage: 收尾 cap24 top2 真实 FMA 对照并确认默认策略
 
 完成项：

@@ -413,6 +413,48 @@ cap24 top2 最终结论：
 - `hybrid`：`16 / 1.0 / 1.0`
 - `high_energy`：`16 / 0.8125 / 1.0`
 - 这个结果比 cap16 更能说明问题：**当前默认策略应明确固定为 `hybrid`**
+
+---
+
+## 11. cap32 top2 对照实验（进行中）
+
+为了确认 cap24 的结论不是偶然，已继续启动更大的真实 FMA top2 对照：
+
+```bash
+cd /workspace/acr-engine
+/usr/local/miniconda3/bin/python scripts/ab_smoke_segmentation.py \
+  --dataset fma \
+  --input-dir data/raw/fma_small_audio \
+  --work-root /tmp/ab_smoke_seg_cap32_top2 \
+  --subset-size 32 \
+  --query-duration 8 \
+  --train-epochs 1 \
+  --batch-size 2 \
+  --device cpu \
+  --strategies hybrid high_energy \
+  --max-test-queries 20 \
+  --output-json /tmp/ab_smoke_seg_cap32_top2/report.json
+```
+
+当前已确认的 fresh evidence：
+
+| 项目 | 状态 |
+|---|---|
+| `subset_size` | `32` |
+| `max_test_queries` | `20` |
+| 首个运行策略 | `hybrid` |
+| 当前阶段 | `run_demo.py build-index --resume --checkpoint-every-refs 100` |
+| `report.json` | 尚未生成 |
+
+恢复检查命令：
+
+```bash
+pgrep -af 'ab_smoke_seg_cap32_top2|external_adapters.py smoke-local fma /tmp/ab_smoke_seg_cap32_top2|evaluate.py --data /tmp/ab_smoke_seg_cap32_top2|run_demo.py build-index --data /tmp/ab_smoke_seg_cap32_top2|train.py --data /tmp/ab_smoke_seg_cap32_top2'
+```
+
+优先等待文件：
+- `/tmp/ab_smoke_seg_cap32_top2/hybrid/fma_reports_smoke/eval.json`
+- `/tmp/ab_smoke_seg_cap32_top2/report.json`
 - `b766c74` Make open-dataset manifests trainable end to end
 - `fa23144` Add a single-page open dataset workflow for training prep
 - `af33be3` Condense docs and add manifest validation before training