jina-v3 + deepseek-chat · k=20
run #6 · 2026-06-19 08:23 · git 0562acf · ab:chunk
Retrieval hit@k
55%
Answer accuracy
71%
Cases
45 (40 grounded)
Jina v3 @ 800 vs row1 Jina@350 — deployed embedder, isolates chunk size (350 was inherited from v1 Berkshire tuning)
Δ vs baseline
Compared to qwen3-embedding:0.6b + deepseek-chat · k=20 (2026-06-18 21:17). A retrieval delta with the same generator isolates the embedder/retrieval change.
- ✓ Retrieval solved here (5): q014, q021, q032, q039, q041
- ⚠ Retrieval regressed (6): q006, q008, q010, q024, q029, q037
- ✓ Answer solved here (1): q032
- ⚠ Answer regressed (1): q029
Failures in this run (23)
| Case | Question | Retrieval | Answer | Symptom |
|---|---|---|---|---|
| q005 | What was Berkshire Hathaway's insurance float at the end of 2023? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q006 | What were Berkshire Hathaway's net earnings attributable to shareholders in 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q007 | What was Berkshire Hathaway's insurance float at the end of 2022? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q008 | What was Berkshire Hathaway's shareholders' equity at the end of 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q010 | What were Berkshire Hathaway's after-tax investment gains in 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q014 | By how much did Berkshire Hathaway's insurance float grow from 2022 to 2023? | PASS | FAIL | retrieved but generated wrong (generation/arithmetic) |
| q021 | By how much did Berkshire Hathaway's insurance float grow from 2021 to 2023? | PASS | FAIL | retrieved but generated wrong (generation/arithmetic) |
| q022 | How much did Berkshire Hathaway's operating earnings grow from 2021 to 2023? | PASS | FAIL | retrieved but generated wrong (generation/arithmetic) |
| q023 | How much did Berkshire Hathaway's operating earnings grow from 2021 to 2022? | PASS | FAIL | retrieved but generated wrong (generation/arithmetic) |
| q024 | How much larger were Berkshire Hathaway's 2023 net earnings than its 2023 operating earnings? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q025 | What was JPMorgan Chase's total net revenue in 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q027 | What were JPMorgan Chase's total assets at the end of 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q028 | What was JPMorgan Chase's CET1 capital ratio in 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q029 | What were JPMorgan Chase's cash dividends declared per share in 2023? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q031 | What was Microsoft's net income in fiscal 2023? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q034 | What were Costco's net sales in fiscal 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q036 | What was Costco's net income in fiscal 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q037 | How much did Costco earn from membership fees in fiscal 2023? | FAIL | PASS | answered without the gold chunk (parametric leak?) |
| q038 | Which company had higher net income in 2023, Microsoft or JPMorgan Chase? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q040 | How much higher was Microsoft's net income than JPMorgan Chase's in 2023? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q041 | How much greater was Costco's total revenue than Microsoft's in 2023? | PASS | FAIL | retrieved but generated wrong (generation/arithmetic) |
| q042 | What was the combined net income of Microsoft and JPMorgan Chase in 2023? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |
| q043 | Which had the highest net income in 2023: JPMorgan Chase, Microsoft, or Costco? | FAIL | FAIL | gold chunk not retrieved → answer ungrounded |