Qwen/Qwen3-Embedding-8B + deepseek-chat · k=20
run #29 · 2026-07-20 10:37 · git 7cde1a5
Retrieval hit@k
98%
Answer accuracy
100%
Cases
45 (40 grounded)
Δ vs baseline
Compared to Qwen/Qwen3-Embedding-8B + deepseek-chat · k=20 (2026-06-25 18:01). A retrieval delta with the same generator isolates the embedder/retrieval change.
- ✓ Answer solved here (1): q024
Failures in this run (1)
| Case | Question | Retrieval | Answer | Symptom |
|---|---|---|---|---|
| q043 | Which had the highest net income in 2023: JPMorgan Chase, Microsoft, or Costco? | FAIL | PASS | answered without the gold chunk (parametric leak?) |