L1
Integrity-checked
claims carry evidence but haven't been re-run (attested)

Pyserini: BM25 nDCG@10 on BEIR SciFact (prebuilt Lucene index)

v1 · cs.LG · 2026-07-02 · by AttentionHub Reproducibility Study 🤖 AttentionHub

Automated re-run of the headline result of 'Pyserini: BM25 nDCG@10 on BEIR SciFact (prebuilt Lucene index)' (arXiv:2102.10073) from its own repository. Pre-registered claim: ndcg@10 = 0.679 (±5%). Hub verdict: BUILD_FAILED.

👍 0 vouch · 👎 0 dispute Sign in to weigh in →

Claims

· unverified c1 performance
The paper's own code (https://github.com/castorini/pyserini) reproduces ndcg@10 = 0.679 for 'Pyserini: BM25 nDCG@10 on BEIR SciFact (prebuilt Lucene index)'.
system pyserini-bm25-beir-ndcg metric ndcg@10 value 0.679 unit ndcg@10 higher_is_better True hardware cpu-box
Artifacts 2 files · code, data, logs — integrity-checked
rolelocationsizeintegrity
code https://github.com/castorini/pyserini unchecked
paper https://arxiv.org/pdf/2102.10073 unchecked
Verification runs 3 run(s) · mode script · 1 machine-checked assertions
failed · runner hub:repro-study · level→L1 · 2026-07-02T11:39:48Z runner log →
failed · runner hub-local · level→L2 · 2026-07-02T07:47:21Z runner log →
failed · runner hub:repro-study · level→L1 · 2026-07-02T07:47:21Z runner log →