Skip to content

Current public HTTP benchmark report

This page is rebuilt from a full 60-sample run whenever the documentation workflow builds main. The published report includes the current commit SHA, fingerprints, coverage, aggregates, group slices, and per-sample evidence.

Interpretation boundary

This is a synthetic, deterministic contract benchmark. It verifies the generic HTTP adapter, metric calculations, citations, provenance, reporting, and CI gates. It does not estimate production RAG quality, model quality, or vendor performance.

Read the benchmark method and limitations or reproduce it locally with the commands on that page.

public-http-retrieval-benchmark

Field Value
ragproof version 0.4.1
timestamp 2026-09-06T07:20:37.184095+00:00
git SHA 3a6b7efe283b5273bb7e58c56057636d8f367bb1
samples 60 / 60
duration 168 ms

Reproducibility

Fingerprint Value
schema_version 2
dataset_sha256 fa1ca93a2b1b067bf7d6544d41ca747c9eda0f3490ec7ad6f9f00430ea329ba9
config_sha256 5f23d44c3a2c338568a752e58348761db3df7232a8fca4330b335e5c2253232d
config_fingerprint_version 2
legacy_config_sha256 9feea0c434df5ec7bc697afd0328fcc3c23a6e4c18a48dc196b5b300dc0debcc
selected_sample_ids_sha256 ff438b0d120d706d00012d4a250f7f54187d1f7ea86cea578489b1df1d9c874d
dataset_manifest {'path': 'examples/dataset.public-http.en.jsonl', 'format': 'jsonl', 'sample_count': 60, 'sha256': 'fa1ca93a2b1b067bf7d6544d41ca747c9eda0f3490ec7ad6f9f00430ea329ba9', 'sample_ids_sha256': 'ff438b0d120d706d00012d4a250f7f54187d1f7ea86cea578489b1df1d9c874d', 'tags': ['chinese', 'citation', 'english', 'hard-negative', 'multihop', 'paraphrase', 'synthetic', 'unanswerable'], 'difficulties': ['hard', 'medium']}
judge_prompt_version builtin-v1
judge_prompt_sha256 None
adapter_type http
group_by ['tags', 'difficulty']

Data Coverage

Field / Metric Available Total Rate
successful_requests 60 60 100.0%
answers 60 60 100.0%
contexts 55 60 91.7%
context_ids 55 60 91.7%
citations 55 60 91.7%

| citation_coverage | 55 | 60 | 91.7% | | citation_precision | 55 | 60 | 91.7% | | citation_recall | 55 | 60 | 91.7% | | citation_span_overlap | 55 | 60 | 91.7% | | citation_validity | 55 | 60 | 91.7% | | claim_support | 55 | 60 | 91.7% | | context_diversity | 55 | 60 | 91.7% | | context_redundancy | 55 | 60 | 91.7% | | context_utilization | 55 | 60 | 91.7% | | duplicate_rate | 60 | 60 | 100.0% | | empty_answer_rate | 60 | 60 | 100.0% | | exact_match | 60 | 60 | 100.0% | | hit_rate@3 | 55 | 60 | 91.7% | | lexical_token_f1 | 60 | 60 | 100.0% | | map@3 | 55 | 60 | 91.7% | | mrr | 55 | 60 | 91.7% | | ndcg@3 | 55 | 60 | 91.7% | | negative_hit_rate | 5 | 60 | 8.3% | | output_token_count | 60 | 60 | 100.0% | | precision@3 | 55 | 60 | 91.7% | | rank_sensitivity@3 | 55 | 60 | 91.7% | | recall@3 | 55 | 60 | 91.7% | | refusal_rate | 60 | 60 | 100.0% | | semantic_similarity | 60 | 60 | 100.0% | | tokens_per_second | 60 | 60 | 100.0% | | unanswerable_correctness | 60 | 60 | 100.0% |

Aggregate Metrics

Metric Value
citation_coverage 1.0000
citation_precision 1.0000
citation_recall 1.0000
citation_span_overlap 1.0000
citation_validity 1.0000
claim_support 1.0000
context_diversity 0.5618
context_redundancy 0.4382
context_utilization 1.0000
duplicate_rate 0.0000
empty_answer_rate 0.0000
exact_match 1.0000
hit_rate@3 1.0000
lexical_token_f1 1.0000
map@3 1.0000
mrr 1.0000
ndcg@3 1.0000
negative_hit_rate 0.0000
output_token_count 6.7167
precision@3 0.3636
rank_sensitivity@3 0.2127
recall@3 1.0000
refusal_rate 0.0000
semantic_similarity 1.0000
tokens_per_second 1897.4923
unanswerable_correctness 1.0000
avg_latency_ms 3.8312
p50_latency_ms 3.9741
p95_latency_ms 4.7786
stream_rate 0.0000
error_rate 0.0000

Groups

tags

Group Metrics
chinese citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.667; context_redundancy=0.333; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=4.000; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1080.265; unanswerable_correctness=1.000; avg_latency_ms=3.812; p50_latency_ms=4.033; p95_latency_ms=4.413; stream_rate=0.000; error_rate=0.000
citation citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.500; context_redundancy=0.500; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=8.000; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=2203.694; unanswerable_correctness=1.000; avg_latency_ms=3.655; p50_latency_ms=3.822; p95_latency_ms=3.894; stream_rate=0.000; error_rate=0.000
english citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.482; context_redundancy=0.518; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=8.000; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=2069.102; unanswerable_correctness=1.000; avg_latency_ms=4.166; p50_latency_ms=4.083; p95_latency_ms=6.244; stream_rate=0.000; error_rate=0.000
hard-negative citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.492; context_redundancy=0.508; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; negative_hit_rate=0.000; output_token_count=8.000; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1914.430; unanswerable_correctness=1.000; avg_latency_ms=4.230; p50_latency_ms=4.109; p95_latency_ms=4.917; stream_rate=0.000; error_rate=0.000
multihop citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.684; context_redundancy=0.316; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=6.000; precision@3=0.667; rank_sensitivity@3=0.167; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1772.290; unanswerable_correctness=1.000; avg_latency_ms=3.570; p50_latency_ms=3.346; p95_latency_ms=4.513; stream_rate=0.000; error_rate=0.000
paraphrase citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.739; context_redundancy=0.261; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=7.000; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1956.425; unanswerable_correctness=1.000; avg_latency_ms=3.662; p50_latency_ms=3.108; p95_latency_ms=4.215; stream_rate=0.000; error_rate=0.000
synthetic citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.562; context_redundancy=0.438; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; negative_hit_rate=0.000; output_token_count=6.717; precision@3=0.364; rank_sensitivity@3=0.213; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1897.492; unanswerable_correctness=1.000; avg_latency_ms=3.831; p50_latency_ms=3.974; p95_latency_ms=4.779; stream_rate=0.000; error_rate=0.000
unanswerable duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; lexical_token_f1=1.000; output_token_count=7.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=3392.097; unanswerable_correctness=1.000; avg_latency_ms=2.250; p50_latency_ms=1.989; p95_latency_ms=3.372; stream_rate=0.000; error_rate=0.000

difficulty

Group Metrics
hard citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.588; context_redundancy=0.412; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; negative_hit_rate=0.000; output_token_count=7.000; precision@3=0.500; rank_sensitivity@3=0.192; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1843.360; unanswerable_correctness=1.000; avg_latency_ms=3.900; p50_latency_ms=4.058; p95_latency_ms=4.917; stream_rate=0.000; error_rate=0.000
medium citation_coverage=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; citation_validity=1.000; claim_support=1.000; context_diversity=0.556; context_redundancy=0.444; context_utilization=1.000; duplicate_rate=0.000; empty_answer_rate=0.000; exact_match=1.000; hit_rate@3=1.000; lexical_token_f1=1.000; map@3=1.000; mrr=1.000; ndcg@3=1.000; output_token_count=6.660; precision@3=0.333; rank_sensitivity@3=0.217; recall@3=1.000; refusal_rate=0.000; semantic_similarity=1.000; tokens_per_second=1908.319; unanswerable_correctness=1.000; avg_latency_ms=3.817; p50_latency_ms=3.913; p95_latency_ms=4.779; stream_rate=0.000; error_rate=0.000

Per-sample Results

ID Question Answer Tags Retrieved IDs Metrics Error
public-001 What approval tier applies to rpcode001 travel reimbursement? rpcode001 travel reimbursement policy requires approval tier 2. english, synthetic doc-001, doc-002, doc-003 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=849.378
public-002 What approval tier applies to rpcode002 travel reimbursement? rpcode002 travel reimbursement policy requires approval tier 3. english, synthetic doc-002, doc-001, doc-003 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1281.322
public-003 What approval tier applies to rpcode003 travel reimbursement? rpcode003 travel reimbursement policy requires approval tier 4. english, synthetic doc-003, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2879.391
public-004 What approval tier applies to rpcode004 travel reimbursement? rpcode004 travel reimbursement policy requires approval tier 1. english, synthetic doc-004, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2313.163
public-005 What approval tier applies to rpcode005 travel reimbursement? rpcode005 travel reimbursement policy requires approval tier 2. english, synthetic doc-005, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1861.099
public-006 What approval tier applies to rpcode006 travel reimbursement? rpcode006 travel reimbursement policy requires approval tier 3. english, synthetic doc-006, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=3585.846
public-007 What approval tier applies to rpcode007 travel reimbursement? rpcode007 travel reimbursement policy requires approval tier 4. english, synthetic doc-007, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1741.248
public-008 What approval tier applies to rpcode008 travel reimbursement? rpcode008 travel reimbursement policy requires approval tier 1. english, synthetic doc-008, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2461.727
public-009 What approval tier applies to rpcode009 travel reimbursement? rpcode009 travel reimbursement policy requires approval tier 2. english, synthetic doc-009, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=2044.688
public-010 What approval tier applies to rpcode010 travel reimbursement? rpcode010 travel reimbursement policy requires approval tier 3. english, synthetic doc-010, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=2013.464
public-011 What approval tier applies to rpcode011 travel reimbursement? rpcode011 travel reimbursement policy requires approval tier 4. english, synthetic doc-011, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1674.140
public-012 What approval tier applies to rpcode012 travel reimbursement? rpcode012 travel reimbursement policy requires approval tier 1. english, synthetic doc-012, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2371.320
public-013 What approval tier applies to rpcode013 travel reimbursement? rpcode013 travel reimbursement policy requires approval tier 2. english, synthetic doc-013, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=2003.210
public-014 What approval tier applies to rpcode014 travel reimbursement? rpcode014 travel reimbursement policy requires approval tier 3. english, synthetic doc-014, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1845.525
public-015 What approval tier applies to rpcode015 travel reimbursement? rpcode015 travel reimbursement policy requires approval tier 4. english, synthetic doc-015, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1755.707
public-016 What approval tier applies to rpcode016 travel reimbursement? rpcode016 travel reimbursement policy requires approval tier 1. english, synthetic doc-016, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2716.929
public-017 What approval tier applies to rpcode017 travel reimbursement? rpcode017 travel reimbursement policy requires approval tier 2. english, synthetic doc-017, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1794.792
public-018 What approval tier applies to rpcode018 travel reimbursement? rpcode018 travel reimbursement policy requires approval tier 3. english, synthetic doc-018, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=2501.733
public-019 What approval tier applies to rpcode019 travel reimbursement? rpcode019 travel reimbursement policy requires approval tier 4. english, synthetic doc-019, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1940.877
public-020 What approval tier applies to rpcode020 travel reimbursement? rpcode020 travel reimbursement policy requires approval tier 1. english, synthetic doc-020, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1907.108
public-021 What approval tier applies to rpcode021 travel reimbursement? rpcode021 travel reimbursement policy requires approval tier 2. english, synthetic doc-021, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=2095.350
public-022 What approval tier applies to rpcode022 travel reimbursement? rpcode022 travel reimbursement policy requires approval tier 3. english, synthetic doc-022, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1873.943
public-023 What approval tier applies to rpcode023 travel reimbursement? rpcode023 travel reimbursement policy requires approval tier 4. english, synthetic doc-023, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1674.933
public-024 What approval tier applies to rpcode024 travel reimbursement? rpcode024 travel reimbursement policy requires approval tier 1. english, synthetic doc-024, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2581.362
public-025 What approval tier applies to rpcode025 travel reimbursement? rpcode025 travel reimbursement policy requires approval tier 2. english, synthetic doc-025, doc-001, doc-002 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1959.285
public-026 rpcode026 的数据保留期限是多少天？ rpcode026 中文知识库规定数据保留期限为 26 天。 chinese, synthetic doc-026, doc-027, doc-028 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=955.031
public-027 rpcode027 的数据保留期限是多少天？ rpcode027 中文知识库规定数据保留期限为 27 天。 chinese, synthetic doc-027, doc-026, doc-028 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=991.131
public-028 rpcode028 的数据保留期限是多少天？ rpcode028 中文知识库规定数据保留期限为 28 天。 chinese, synthetic doc-028, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1006.506
public-029 rpcode029 的数据保留期限是多少天？ rpcode029 中文知识库规定数据保留期限为 29 天。 chinese, synthetic doc-029, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1044.541
public-030 rpcode030 的数据保留期限是多少天？ rpcode030 中文知识库规定数据保留期限为 30 天。 chinese, synthetic doc-030, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1023.565
public-031 rpcode031 的数据保留期限是多少天？ rpcode031 中文知识库规定数据保留期限为 31 天。 chinese, synthetic doc-031, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1460.458
public-032 rpcode032 的数据保留期限是多少天？ rpcode032 中文知识库规定数据保留期限为 32 天。 chinese, synthetic doc-032, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1207.209
public-033 rpcode033 的数据保留期限是多少天？ rpcode033 中文知识库规定数据保留期限为 33 天。 chinese, synthetic doc-033, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=906.453
public-034 rpcode034 的数据保留期限是多少天？ rpcode034 中文知识库规定数据保留期限为 34 天。 chinese, synthetic doc-034, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1643.285
public-035 rpcode035 的数据保留期限是多少天？ rpcode035 中文知识库规定数据保留期限为 35 天。 chinese, synthetic doc-035, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=991.878
public-036 rpcode036 的数据保留期限是多少天？ rpcode036 中文知识库规定数据保留期限为 36 天。 chinese, synthetic doc-036, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=981.577
public-037 rpcode037 的数据保留期限是多少天？ rpcode037 中文知识库规定数据保留期限为 37 天。 chinese, synthetic doc-037, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=870.949
public-038 rpcode038 的数据保留期限是多少天？ rpcode038 中文知识库规定数据保留期限为 38 天。 chinese, synthetic doc-038, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=1188.679
public-039 rpcode039 的数据保留期限是多少天？ rpcode039 中文知识库规定数据保留期限为 39 天。 chinese, synthetic doc-039, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=976.088
public-040 rpcode040 的数据保留期限是多少天？ rpcode040 中文知识库规定数据保留期限为 40 天。 chinese, synthetic doc-040, doc-026, doc-027 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.333; context_diversity=0.667; claim_support=1.000; output_token_count=4.000; tokens_per_second=956.626
public-041 UNANSWERABLE rpnone041 asks for information absent from the corpus. I cannot answer from the provided context. synthetic, unanswerable `` recall@3=N/A; precision@3=N/A; ndcg@3=N/A; map@3=N/A; hit_rate@3=N/A; rank_sensitivity@3=N/A; mrr=N/A; duplicate_rate=0.000; citation_coverage=N/A; citation_validity=N/A; citation_precision=N/A; citation_recall=N/A; citation_span_overlap=N/A; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=N/A; context_redundancy=N/A; context_diversity=N/A; claim_support=N/A; output_token_count=7.000; tokens_per_second=2705.587
public-042 UNANSWERABLE rpnone042 asks for information absent from the corpus. I cannot answer from the provided context. synthetic, unanswerable `` recall@3=N/A; precision@3=N/A; ndcg@3=N/A; map@3=N/A; hit_rate@3=N/A; rank_sensitivity@3=N/A; mrr=N/A; duplicate_rate=0.000; citation_coverage=N/A; citation_validity=N/A; citation_precision=N/A; citation_recall=N/A; citation_span_overlap=N/A; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=N/A; context_redundancy=N/A; context_diversity=N/A; claim_support=N/A; output_token_count=7.000; tokens_per_second=4948.129
public-043 UNANSWERABLE rpnone043 asks for information absent from the corpus. I cannot answer from the provided context. synthetic, unanswerable `` recall@3=N/A; precision@3=N/A; ndcg@3=N/A; map@3=N/A; hit_rate@3=N/A; rank_sensitivity@3=N/A; mrr=N/A; duplicate_rate=0.000; citation_coverage=N/A; citation_validity=N/A; citation_precision=N/A; citation_recall=N/A; citation_span_overlap=N/A; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=N/A; context_redundancy=N/A; context_diversity=N/A; claim_support=N/A; output_token_count=7.000; tokens_per_second=3519.211
public-044 UNANSWERABLE rpnone044 asks for information absent from the corpus. I cannot answer from the provided context. synthetic, unanswerable `` recall@3=N/A; precision@3=N/A; ndcg@3=N/A; map@3=N/A; hit_rate@3=N/A; rank_sensitivity@3=N/A; mrr=N/A; duplicate_rate=0.000; citation_coverage=N/A; citation_validity=N/A; citation_precision=N/A; citation_recall=N/A; citation_span_overlap=N/A; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=N/A; context_redundancy=N/A; context_diversity=N/A; claim_support=N/A; output_token_count=7.000; tokens_per_second=3711.579
public-045 UNANSWERABLE rpnone045 asks for information absent from the corpus. I cannot answer from the provided context. synthetic, unanswerable `` recall@3=N/A; precision@3=N/A; ndcg@3=N/A; map@3=N/A; hit_rate@3=N/A; rank_sensitivity@3=N/A; mrr=N/A; duplicate_rate=0.000; citation_coverage=N/A; citation_validity=N/A; citation_precision=N/A; citation_recall=N/A; citation_span_overlap=N/A; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=N/A; context_redundancy=N/A; context_diversity=N/A; claim_support=N/A; output_token_count=7.000; tokens_per_second=2075.980
public-046 Combine the retention and access phases for rpcode046. rpcode046 retention phase is 46 days. multihop, synthetic doc-046-a, doc-046-b, doc-047-a recall@3=1.000; precision@3=0.667; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.167; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.316; context_diversity=0.684; claim_support=1.000; output_token_count=6.000; tokens_per_second=2517.641
public-047 Combine the retention and access phases for rpcode047. rpcode047 retention phase is 47 days. multihop, synthetic doc-047-a, doc-047-b, doc-046-a recall@3=1.000; precision@3=0.667; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.167; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.316; context_diversity=0.684; claim_support=1.000; output_token_count=6.000; tokens_per_second=1380.484
public-048 Combine the retention and access phases for rpcode048. rpcode048 retention phase is 48 days. multihop, synthetic doc-048-a, doc-048-b, doc-046-a recall@3=1.000; precision@3=0.667; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.167; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.316; context_diversity=0.684; claim_support=1.000; output_token_count=6.000; tokens_per_second=1793.184
public-049 Combine the retention and access phases for rpcode049. rpcode049 retention phase is 49 days. multihop, synthetic doc-049-a, doc-049-b, doc-046-a recall@3=1.000; precision@3=0.667; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.167; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.316; context_diversity=0.684; claim_support=1.000; output_token_count=6.000; tokens_per_second=1329.576
public-050 Combine the retention and access phases for rpcode050. rpcode050 retention phase is 50 days. multihop, synthetic doc-050-a, doc-050-b, doc-046-a recall@3=1.000; precision@3=0.667; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.167; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.316; context_diversity=0.684; claim_support=1.000; output_token_count=6.000; tokens_per_second=1840.567
public-051 Which signed evidence level does rpcode051 verification warranty require? rpcode051 verification warranty requires signed evidence level 4. hard-negative, synthetic doc-051, doc-052, doc-053 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1971.638; negative_hit_rate=0.000
public-052 Which signed evidence level does rpcode052 verification warranty require? rpcode052 verification warranty requires signed evidence level 1. hard-negative, synthetic doc-052, doc-051, doc-053 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1946.732; negative_hit_rate=0.000
public-053 Which signed evidence level does rpcode053 verification warranty require? rpcode053 verification warranty requires signed evidence level 2. hard-negative, synthetic doc-053, doc-051, doc-052 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=1626.887; negative_hit_rate=0.000
public-054 Which signed evidence level does rpcode054 verification warranty require? rpcode054 verification warranty requires signed evidence level 3. hard-negative, synthetic doc-054, doc-051, doc-052 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2258.137; negative_hit_rate=0.000
public-055 Which signed evidence level does rpcode055 verification warranty require? rpcode055 verification warranty requires signed evidence level 4. hard-negative, synthetic doc-055, doc-051, doc-052 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.542; context_diversity=0.458; claim_support=1.000; output_token_count=8.000; tokens_per_second=1768.759; negative_hit_rate=0.000
public-056 Cite the evidence record for rpcode056. evidence record rpcode056 citation rule points to evidence record 56. citation, synthetic doc-056, doc-057, doc-058 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2463.716
public-057 Cite the evidence record for rpcode057. evidence record rpcode057 citation rule points to evidence record 57. citation, synthetic doc-057, doc-056, doc-058 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2093.083
public-058 Cite the evidence record for rpcode058. evidence record rpcode058 citation rule points to evidence record 58. citation, synthetic doc-058, doc-056, doc-057 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.500; context_diversity=0.500; claim_support=1.000; output_token_count=8.000; tokens_per_second=2054.284
public-059 In different words, how quickly should rpcode059 return to service? rpcode059 service restoration objective is 59 minutes. paraphrase, synthetic doc-059, doc-056, doc-057 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.261; context_diversity=0.739; claim_support=1.000; output_token_count=7.000; tokens_per_second=1660.679
public-060 In different words, how quickly should rpcode060 return to service? rpcode060 service restoration objective is 60 minutes. paraphrase, synthetic doc-060, doc-056, doc-057 recall@3=1.000; precision@3=0.333; ndcg@3=1.000; map@3=1.000; hit_rate@3=1.000; rank_sensitivity@3=0.217; mrr=1.000; duplicate_rate=0.000; citation_coverage=1.000; citation_validity=1.000; citation_precision=1.000; citation_recall=1.000; citation_span_overlap=1.000; exact_match=1.000; lexical_token_f1=1.000; semantic_similarity=1.000; empty_answer_rate=0.000; refusal_rate=0.000; unanswerable_correctness=1.000; context_utilization=1.000; context_redundancy=0.261; context_diversity=0.739; claim_support=1.000; output_token_count=7.000; tokens_per_second=2252.172