兩項目評估報告
- EvalRun: #277（suite #126
auto-319f5c68-r1-053959578）
- Target: #1 production-baseline
- Cases: 20（N=1, attempts=20/20）
- Generated: 2026-05-16T14:59:13+08:00
本報告含兩個獨立評估項目。項目一 評估 bot 進入「知識與產品查詢」fallback
後的撈取 + 回答能力；項目二 評估 bot 跨全 enabled scenarios 的 routing →
tool → answer 三段 funnel 健康度。
項目一: 知識庫精準度
評估範圍 (prerequisite: 進對 scenario)
- 進對 scenario 的 attempts: 4 / 20 (20.0%)
- 只對這些 attempts 算 retrieval + answer
- 沒進對 scenario 的 attempts → 排除（routing 失敗不該污染 KB 訊號）
Per-scenario 評估
| Scenario |
qualifying / total attempts |
retrieval_relevance |
answer_correctness |
| 知識與產品查詢 |
4 / 4 |
❌ 0.0% |
— |
待修方向（worst-3）
- 知識與產品查詢 × retrieval_relevance = 0.0% — bot 進對 scenario 但沒呼叫 search tool — 流程 / prompt 問題（不是 KB content gap）
項目二: 情境調用與完成
整體 funnel（全 scenarios 加總）
| Stage |
Pass count |
% of total |
| Total attempts |
16 |
100.0% |
| Step 1: scenario_routing |
11 |
68.8% |
| Step 2: + tool_calling |
11 |
68.8% |
| Step 3: + answer |
8 |
50.0% |
Per-scenario funnel
| Scenario (n_attempts) |
Step 1 (routing) |
Step 2 (tool) |
Step 3 (answer) |
end-to-end |
| early_return (4) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
| 知識與產品查詢 (11) |
✅ 10/11 (90.9%) |
✅ 10/11 (90.9%) |
⚪ 7/11 (63.6%) |
⚪ 7/11 (63.6%) |
| 轉接真人客服 (1) |
✅ 1/1 (100.0%) |
✅ 1/1 (100.0%) |
✅ 1/1 (100.0%) |
✅ 1/1 (100.0%) |
Drop-off 最大的 5 個 scenario
- early_return drop at step1 (-100.0pp)
- 知識與產品查詢 drop at step3 (-27.3pp)
- 轉接真人客服 drop at step1 (-0.0pp)
Audit
Reproduce combined report: bin/rails runner "puts Eval::EvaluationReport.call(run: EvalRun.find(277))"
Or fetch each item separately:
bin/rails runner "puts Eval::KbAccuracyReport.call(run: EvalRun.find(277))"bin/rails runner "puts Eval::ScenarioFunnelReport.call(run: EvalRun.find(277))"
Per-Scenario × Per-Dim — Run #277
Suite: test (bulk R1) · scenarios: 3 · dims: 5 · populated cells: 12/15
| Scenario |
Scenario |
Tool |
Retrieval |
Faith |
AnsQ |
| early_return |
0.0% [0.0–0.0] (n=4) |
— |
100.0% [100.0–100.0] (n=4) |
0.0% (n=1) |
97.5% [92.5–100.0] (n=4) |
| 知識與產品查詢 |
93.3% [80.0–100.0] (n=15) |
93.3% [80.0–100.0] (n=15) |
92.9% [82.1–100.0] (n=14) |
57.1% [28.6–85.7] (n=7) |
84.7% [70.7–95.8] (n=15) |
| 轉接真人客服 |
100.0% (n=1) |
— |
— |
100.0% (n=1) |
100.0% (n=1) |
Worst-3 cells (lowest primary score)
- early_return × Scenario · 0.0% (n=4) · lowest sub_metric:
scenario_precision
- early_return × Faith · 0.0% (n=1) · lowest sub_metric:
hallucination_rate
- 知識與產品查詢 × Faith · 57.1% (n=7) · lowest sub_metric:
rule_compliance