Run 42d58a0 · Own gold set (private corpus)

Run 42d58a0 Own gold set (private corpus), recorded 25 June 2026, 16:53:03 UTC.

dirty tree · not cited
determinism ok
private corpus · not reproducible

Benchmark metrics
Metric	Value	Definition
Recall@5	0.9554	Mean over questions of \|evidence ∩ top-k\| / \|evidence\|.
nDCG@5	0.9329	With binary relevance.
MRR	0.9389	Mean of 1/rank of the first evidence session.
Precision@5	0.2143	Mean over questions of \|evidence ∩ top-k\| / k. It can't reach 1 when a question has fewer than k evidence sessions, so it is shown on receipts only.

Conditions
Queries	56
k	5
Strategy	bm25
Gold path	benchmarks/recall-gold.json
Tests ran	332
Tests failed	0
Environment	Hardware, OS and Python version: not recorded for this run.
Ledger row	87

Raw row

{"ablation": {"baseline": "tag", "delta": {"context_tokens_mean": 2547.3035714285716, "latency_ms_mean": 8.596950875861305, "latency_ms_p50": 8.544041513232514, "mrr": 0.41371173469387745, "ndcg_at_k": 0.40848740988791643, "precision_at_k": 0.08214285714285713, "recall_at_k": 0.3392857142857143}, "treatment": "bm25"}, "checks": {"determinism_ok": true, "index_validate_ok": true, "test_pass_rate": 1.0, "tests_failed": 0, "tests_ran": 332}, "commit": "42d58a0", "committed": "2026-06-25T15:34:24Z", "dirty": true, "gold_path": "benchmarks/recall-gold.json", "metrics": {"context_tokens_mean": 3947.9821428571427, "latency_ms_mean": 8.60752301845683, "latency_ms_p50": 8.553874504286796, "mrr": 0.938860544217687, "ndcg_at_k": 0.9329032950598203, "precision_at_k": 0.21428571428571427, "recall_at_k": 0.9553571428571429}, "mode": "ablate", "num_queries": 56, "reliability": 98.21, "run": "2026-06-25T16:53:03Z", "schema": "agentware-bench/v1", "strategy": "bm25", "subject": "feat(benchmark): LongMemEval scorer + Mode A baseline + public repro docs", "suite": null, "top_k": 5}

reliability is an internal composite health score (0–100), not a benchmark metric.