run_2026-08-03_v3-3-release
support-assistant v3.3 · release
Claude Sonnet 5.5 · bge-m3 · hybrid (BM25 + dense, RRF) · support-prompt v6 on support-golden-v4 v4.6
Prompt v6 and hybrid retrieval. +3.7 pts pass rate vs v3.2.
Pass rate
81.2%
Faithfulness
0.87
Answer correctness
0.84
Context precision
0.74
Context recall
0.79
Latency p95
3.7s
Failures by category
233 failed samples · largest share: unsupported
| Category | Share | vs reference |
|---|---|---|
| Unsupported claims | 25.8% | +1.7 pts |
| Incorrect factual answers | 20.2% | +1.2 pts |
| Missing context | 21.0% | −1.9 pts |
| Retrieval mismatch | 21.0% | −3.0 pts |
| Instruction-following failures | 9.0% | +1.8 pts |
| Other | 3.0% | +0.1 pts |
Run details
- Dataset
- support-golden-v4 v4.6 · 1,240 items
- Generator
- Claude Sonnet 5.5
- Retriever
- bge-m3 · hybrid (BM25 + dense, RRF)
- Prompt
- support-prompt v6
- Judge
- LLM judge · Claude Opus 5.5 · rubric v3
- Started
- Aug 3, 13:00 UTC
- Duration
- 2h 46m
- Triggered by
- Maya Collins
- Tokens / request
- 3,980 in · 221 out
- Compared with
- support-assistant v3.2 · release
- release
- regression-gate
All metrics
Against support-assistant v3.2 · release
| Metric | Group | This run | Reference | Change |
|---|---|---|---|---|
| Pass rateShare of samples the judge marked as passing all rubric checks. | Outcome | 81.2% | 77.5% | +3.7 pts |
| FaithfulnessShare of answer claims supported by the retrieved context. | Generation | 0.87 | 0.85 | +0.02 |
| Answer correctnessAgreement with the gold answer on facts, not style. | Generation | 0.84 | 0.81 | +0.03 |
| Instruction adherenceRespect for tone, format, length, language and policy constraints. | Generation | 0.89 | 0.87 | +0.02 |
| Context precisionRank-weighted precision of relevant chunks passed to the generator. | Retrieval | 0.74 | 0.70 | +0.04 |
| Context recallShare of gold-answer statements supported by the retrieved context. | Retrieval | 0.79 | 0.75 | +0.04 |
| Recall@5Share of gold passages found in the top 5 retrieved chunks. | Retrieval | 0.83 | 0.78 | +0.05 |
| MRRMean reciprocal rank of the first relevant chunk. | Retrieval | 0.69 | 0.64 | +0.05 |
| NDCG@10Graded ranking quality of the top 10, normalized by the ideal order. | Retrieval | 0.76 | 0.72 | +0.04 |
| Latency p50Median end-to-end time to final token. | Operational | 1.8s | 1.7s | +0.1s |
| Latency p9595th percentile end-to-end time to final token. | Operational | 3.7s | 3.5s | +0.2s |
| Cost per requestModel, embedding and reranking cost per answered request (USD). | Operational | $0.0046 | $0.0044 | +0.0002 |
Nightly trend
Nightly runs on the support-golden-v4 slice, Jul 11 to Oct 8. Dashed line marks this run.
- Pass rate · nightly
- This run · 81.2%
Across datasets
support-assistant v3.3 on every dataset in the project. Rows marked “matrix” come from the comparison matrix rather than a dedicated run.
| Dataset | Items | Pass rate | Faith. | Ctx recall | Recall@5 | p95 | Source |
|---|---|---|---|---|---|---|---|
| support-golden-v4this run | 1,240 | 81.2% | 0.87 | 0.79 | 0.83 | 3.7s | This run |
| billing-edge-cases | 312 | 76.0% | 0.83 | 0.77 | 0.81 | 3.9s | Run · Aug 20 |
| roaming-and-travel | 188 | 75.5% | 0.84 | 0.74 | 0.78 | 3.8s | matrix |
| device-troubleshooting | 426 | 78.9% | 0.85 | 0.78 | 0.82 | 4.4s | matrix |
| account-security | 154 | 82.5% | 0.88 | 0.80 | 0.84 | 3.6s | matrix |
| locale-formatting | 96 | 75.0% | 0.82 | 0.75 | 0.79 | 4.0s | matrix |
| synthetic-hard-negatives | 540 | 70.7% | 0.82 | 0.68 | 0.72 | 3.9s | matrix |
Traced samples
17 samples with full traces (4 failing, 13 passing), stratified by domain. Open a row to inspect retrieval and the judge's reasoning.
Is there a restocking fee if I return an opened phone? Returns & warranty FailMissingjudge 0.13 |
Can I pay my bill with a credit card from another person? Billing & payments FailIncorrectjudge 0.58 |
Does the Fall 2026 upgrade offer work with a family plan discount? Promotions Passjudge 0.93 |
What is the late fee if I pay my bill after the due date? Billing & payments Passjudge 0.95 |
How long does it take to port my number to Harbor? Number porting Passjudge 0.90 |
How do I get a copy of the personal data Harbor holds about me? Data privacy requests Passjudge 0.89 |
My phone shows 4G but mobile data does not work after the iOS update. Device troubleshooting Passjudge 0.86 |
Why was I charged twice for my September bill? Billing & payments Passjudge 0.93 |
How much data can I use in the EU on the Unlimited Plus plan? Roaming & travel Passjudge 0.91 |
If I upgrade my plan mid-month, how is my next bill calculated? Plans & upgrades Passjudge 0.91 |