兩項目評估報告
- EvalRun: #246(suite #135
auto-64c058af-r1-053959682)
- Target: #2 production-canary
- Cases: 15(N=1, attempts=15/15)
- Generated: 2026-05-16T13:51:02+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 / 15 (26.7%)
- 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 |
⚪ 51.2% |
✅ 100.0% |
待修方向(worst-3)
- 知識與產品查詢 × retrieval_relevance = 51.2% — retrieval 略低 — 看 LLM relevance vs topic_match 哪邊弱
- 知識與產品查詢 × answer_correctness = 100.0% — 回答內容有問題 — 細看 hallucination_rate vs answer_quality 哪邊低
項目二: 情境調用與完成
整體 funnel(全 scenarios 加總)
| Stage |
Pass count |
% of total |
| Total attempts |
11 |
100.0% |
| Step 1: scenario_routing |
0 |
0.0% |
| Step 2: + tool_calling |
0 |
0.0% |
| Step 3: + answer |
0 |
0.0% |
Per-scenario funnel
| Scenario (n_attempts) |
Step 1 (routing) |
Step 2 (tool) |
Step 3 (answer) |
end-to-end |
| FAQ查詢 (4) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
❌ 0/4 (0.0%) |
| early_return (1) |
❌ 0/1 (0.0%) |
❌ 0/1 (0.0%) |
❌ 0/1 (0.0%) |
❌ 0/1 (0.0%) |
| 產品查詢 (3) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
| 訂單查詢 (3) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
❌ 0/3 (0.0%) |
Drop-off 最大的 5 個 scenario
- FAQ查詢 drop at step1 (-100.0pp)
- early_return drop at step1 (-100.0pp)
- 產品查詢 drop at step1 (-100.0pp)
- 訂單查詢 drop at step1 (-100.0pp)
Audit
Reproduce combined report: bin/rails runner "puts Eval::EvaluationReport.call(run: EvalRun.find(246))"
Or fetch each item separately:
bin/rails runner "puts Eval::KbAccuracyReport.call(run: EvalRun.find(246))"
bin/rails runner "puts Eval::ScenarioFunnelReport.call(run: EvalRun.find(246))"
Per-Scenario × Per-Dim — Run #246
Suite: WEEX Agent (bulk R1) · scenarios: 5 · dims: 5 · populated cells: 21/25
| Scenario |
Scenario |
Tool |
Retrieval |
Faith |
AnsQ |
| FAQ查詢 |
0.0% [0.0–0.0] (n=4) |
0.0% [0.0–0.0] (n=4) |
25.0% [0.0–75.0] (n=4) |
33.3% [0.0–75.0] (n=4) |
63.3% [38.3–88.3] (n=4) |
| early_return |
0.0% (n=1) |
0.0% (n=1) |
— |
— |
100.0% (n=1) |
| 產品查詢 |
0.0% [0.0–0.0] (n=3) |
0.0% [0.0–0.0] (n=3) |
— |
0.0% [0.0–0.0] (n=3) |
26.7% [26.7–26.7] (n=3) |
| 知識與產品查詢 |
100.0% [100.0–100.0] (n=4) |
0.0% [0.0–0.0] (n=4) |
100.0% [100.0–100.0] (n=4) |
100.0% [100.0–100.0] (n=4) |
82.5% [65.0–95.8] (n=4) |
| 訂單查詢 |
66.7% [0.0–100.0] (n=3) |
0.0% [0.0–0.0] (n=3) |
— |
100.0% [100.0–100.0] (n=3) |
44.4% [26.7–73.3] (n=3) |
Worst-3 cells (lowest primary score)
- 訂單查詢 × Tool · 0.0% (n=3) · lowest sub_metric:
tools_precision
- 知識與產品查詢 × Tool · 0.0% (n=4) · lowest sub_metric:
tools_precision
- 產品查詢 × Faith · 0.0% (n=3) · lowest sub_metric:
hallucination_rate